The Proceedings of the 5th International Conference on Energy Storage and Intelligent Vehicles (ICEIV 2022)

The capacity fading of lithium iron phosphate batteries is related to its internal temperature and the growth of solid electrolyte (SEI). It is an effective way by controlling its internal temperature to mitigate capacity fading. This paper discusses the impact of pre-cooling and resting time on capacity fading and the growth of SEI. Results showed that the battery capacity increased and the thickness of SEI decreased if the pre-cooling was employed. Compared to 25 °C of ambient temperature, the thickness of SEI under 5 °C of pre-cooling temperature decreased by 404 nm, 386 nm, and 502 nm for 2C, 3C, and 5C discharge rate, respectively. The internal temperature of battery could be better cooled and therefore capacity increased with the increase of resting time. At 15 °C of pre-cooling temperature, the capacity increased by 3.8% if the resting time increased from 600 s to 2400 s. Therefore, the pre-cooling method could effectively mitigate capacity fading. The conclusion obtained in this paper could provide guidance for battery thermal management.

The high frequency magnetic field area in the radio energy transmission system (WPT) is an important medium for electric energy transmission. As an open structure, the high frequency magnetic field often inevitably falls into metal foreign matters, leading to a decline in the transmission efficiency of the system, and the heating of the metal itself will also cause potential safety hazards. In this paper, based on the loose coupling model, the detection sensitivity formula of the detection coil method is established. On this basis, combined with the magnetic field characteristics of the square transmission coil surface, a long rectangular interconnection detection coil group and its layout are proposed. Compared with the traditional small coil independent detection method, this method has the advantages of less detection channels, high detection sensitivity, simple control and less blind area. Through electromagnetic field simulation, the advantages of the proposed structure are proved, and experimental verification is carried out on the WPT platform of the square transmission coil to achieve effective detection of ferrite, copper sheet, iron sheet and aluminum alloy sheet.

High temperature failure of truck brake is one of the main causes of truck accidents, which is an urgent problem to be solved for vehicle safety. Especially in long downhill, the problem is more pronounced due to the frequent braking. In this paper, a brake control strategy is proposed to be applied in hybrid electric trucks. Firstly, the model considering hybrid powertrain and brake temperature rise is established. Then, an optimal method based on Pontryagin's minimum principle (PMP) is designed. To improve the adaptability of the proposed strategy to the road, a self-correcting law is designed adjust the costates in the optimization method based on road information and vehicle state. Finally, four test roads were fitted by reference to several typical highways with long downhills in China. The results that the proposed can significantly reduce the maximum braking temperature with a rule strategy in fitting long downhill. It is also found that the continuous long downhill is more disadvantageous to the brake than the staged downhill.

The Paper presents verification of induction motor control and its characteristics using indirect field-oriented control (IFOC) through a Typhoon Hardware-in-loop (HIL) simulation package with HIL604 configuration for both testbench control and real-time simulation. With field-oriented control, the angle and the magnitude of each phase’s voltage and current are controlled. Typically, torque control is accomplished by modulating the armature with a constant field current. The Field Weakening is used to boost the speed beyond its base value. To decouple the motor torque and flux, a d-q reference frame locked to the rotor flux vector is employed. Therefore, they can be independently controlled by the stator's quadrature-axis current and direct-axis current, respectively. Using the PI controller and SVPWM approach, satisfactory performance is realized. The rotor flux and magnetizing current are obtained using the proposed control strategy. To control the physical testbench, real-time control signals are amplified using PONOVO voltage and current amplifiers, which also make the output signal from the physical motor return at analog input ports of Typhoon. A dynamometer is used to measure the speed and torque of the induction motor. Software and hardware results are compared, and finally, efficiency is calculated.

Estimating the residual capacity of retired batteries (RCRB) is a critical component of second-use applications (SUAs). This paper provides a hybrid model that combines a mechanism and a data-driven approach (MDA) to increase the accuracy of battery residual capacity estimation. First, the Levenberg Marquardt algorithm (LMA) is utilized to extract three health indicators (HI) that directly characterize the mechanism of capacity loss (MCL) from the constant current (CC) charging voltage curve. Second, the support vector regression optimized by the improved whale optimization algorithm (IWOA-SVR) is established, where the IWOA algorithm is constructed by an adaptive weight embedded in the WOA algorithm, which can prevent the local optimal value of WOA algorithm. Finally, 500 retired LiFePO4 (LFP) batteries were used to validate the effectiveness of the proposed method, the results reveal that when only the first 10% of the data is utilized, the root mean square error (RMSE) is 2.65% and the mean absolute error (MAE) is 1.82%. An accurate hybrid estimation model for RCRB can reduce the cost and time required for SUAs.

Safety is a prerequisite for promoting and applying battery energy storage stations (BESS). This paper develops a Li-ion battery BESS full-time safety protection system based on digital twin technology. Firstly, from the source of safety risk of BESS, the multi-physical characteristics of “electrical-gas-sound-image” in the thermal runaway (TR) process are integrated, and a multi-level early warning protection method for BESS is proposed. It can realize real-time sensing and early warning of multi-time scale safety status from regular operation, micro-overcharge, and TR to fire. Based on the digital twin technology, the core features of the BESS digital twin are described in six aspects: the accurate mapping of virtual reality, real-time sensing, and edge processing, and the overall design framework of the digital twin BESS safety protection system is proposed. The digital twin safety protection system can fully use BESS's massive operation data, improve BESS's safety coefficient and uncover potential failure risks, providing a new idea for the digitalization and intelligence of BESS operation supervision and safety production.

The electrical fault diagnosis and state estimation are curial tasks for lithium-ion battery pack based electric drive system. The electrical fault of batteries will lead to abnormal state estimation. Considering the computational cost and limited memory sources of the on-board battery management system, a framework of online electrical fault diagnosis and low-cost state estimation for the lithium-ion battery pack is proposed. Firstly, an available capacity based macro-selection method is carried out to select a “representative cell” of the battery pack periodically. Secondly, the correlation coefficient between “representative cell” and non-representative cells is calculated based on an optimal moving window in dual time-scale, then the electrical fault can be compensated based on the correlation coefficient in time. Finally, the Kalman filter framework is adopted for robust closed-loop state estimation for the lithium-ion battery pack, in which the Gaussian process regression is developed for measurement equation and Ampere hour counting is established for state equation, then the low-cost state estimation can be achieved. Experimental tests on battery packs were carried out under several dynamic load conditions considering user habits. The validation results demonstrate the robustness and accuracy of the proposed approach even there exists multiple disturbances.

The capacity prediction of lithium-ion battery (LIB) plays a very important role in health management and the prediction of the performance degradation degree for battery. Accurate prediction of capacity can guide battery replacement and maintenance, and ensure the security and stability of battery. In this paper, based on the hybrid grey wolf optimizer-support vector regression (HGWO-SVR) algorithm, the capacity of LIB is predicted, and the remaining useful life is represented by the capacity of LIB. Differential evolution algorithm optimizes gray wolf algorithm to avoid it falling into local optimal solution. The hybrid gray wolf algorithm optimizes parameters of the support vector regression algorithm to improve prediction accuracy. The proposed method is simulated and verified by using NASA's LIB capacity database. By comparing different optimization algorithms and different neural networks, it is verified that the proposed algorithm can forecast the LIB’s capacity more accurately.

The recognition of life cycle carbon footprint (LCCF) of power transmission equipment is of great significance to the de-carbonization of power systems. Based on the technical features of power transmission equipment, the concept of LCCF is proposed in this paper, which includes fixed carbon emissions and variable carbon emissions. Firstly, to evaluate the LCCF of power transmission equipment, the life cycle of transmission equipment is divided into five stages: raw materials, manufacturing and assembly, transportation, operation, and maintenance, decommissioning and scrapping according to the time series, and the carbon emissions in all stages of the life cycle are calculated in combination with various carbon emission factors and quality criteria. Then, based on the concept of equal annual value, a conversion model for carbon footprint is proposed to realize the comprehensive evaluation of LCCF. Finally, the correctness and effectiveness of the proposed method are verified by simulation analysis. The results show that fixed carbon emissions account for a certain proportion of LCCF. To achieve the double carbon goal on schedule, it is necessary to study the LCCF of power transmission equipment and make independent contributions to accelerating the process of de-carbonization.

The development of energy-efficient battery thermal management technology is of great significance for lithium-ion batteries. In this paper, a Tesla valve type channel cold plate was designed for square batteries, also liquid cooling experimental studies were carried out to verify the optimized cold plate parameters. The maximum error between liquid cooling simulation and experiment under the optimal configuration did not exceed 1.25 °C. The experimental analysis found that when the inlet flow rate exceeded 398mL/min, the improvement of battery cooling effect and temperature uniformity gradually tended to saturate. The coolant inlet temperature was too high or too low would cause the unbalanced performance of the cold plate.

With the increasing penetration of wind power, how to accurately analyze characteristics of wind power becomes significant to the safe and stable operation of power system. This paper proposes a data-driven based wind power characteristic analysis and climbing identification method. Wind power climbing threshold is extracted firstly to construct climbing event dataset based on k-means clustering. 2D convolutional neural network-based climbing identification method is then proposed, with network parameters trained by transforming 1-dimensional wind power output records into a 2-dimensional matrix to identify future climbing events. Test results on practical wind farm show that the proposed method can effectively analyze characteristics of wind power, which has better climbing identification accuracy compared with traditional methods.

The accurate prediction of remaining useful life (RUL) is vital to improve the safety and reliability of lithium-ion battery power systems. However, owing to the effects of state (work and storage) switching and retention time randomness, it is difficult to accurately predict the RUL of lithium-ion batteries in real time. This study uses a nonlinear-drift-driven Wiener process to describe the dynamic degradation paths of lithium-ion batteries, and utilizes the Markov chain to establish a switching model and to predict the future state-switching probability. The analytical distribution of the first arrival time of the lithium-ion battery failure threshold is derived, and the model posterior parameters are updated using the Bayesian strategy. Finally, the model is verified using a lithium-ion battery dataset from the National Aeronautics and Space Administration. The RUL prediction model that considers state switching is superior to models that do not consider state switching. Moreover, the model shows a relative error of only 5.9388%. Overall, this study provides a theoretical basis for the research and development of lithium-ion battery RUL prediction and health management systems.

Valve-regulated lead-acid batteries are the preferred power supply equipment in substation DC power supplies due to their advantages of small self-discharge, airtightness and easy maintenance. SOH is an important parameter that reflects the current capacity of a battery. In recent years, there have been a variety of algorithms to study SOH state estimation. However, considering its accuracy, engineering application and other factors, it is difficult to meet accurate online measurement. In this paper, the differences in electrical parameters of 2 V 500 Ah VRLA batteries under different health states are analyzed, and the voltage drop value of 0–150 s in the fully charged state and the AC impedance value at 50 Hz frequency are used to estimate the battery. Under the premise of not applying any algorithm optimization and data screening, two electrical parameters are used to estimate the state of the battery. The results show that using the voltage drop and the AC impedance value in the fully charged state has a high accuracy in estimating the state of the battery. This shows a good value for engineering application.

Traction transformer is the core component of EMU traction system. It can convert high-voltage electric energy obtained from the catenary into suitable voltage level for the input of traction container. Due to abundant high order harmonics produced by the four-quadrant rectifier of the traction container, secondary windings of the traction transformer generate ferromagnetic noise which is uncomfortable to human ears. Guaranteeing traction performance and minimizing the ferromagnetic noise by traction transformer at the same time is a difficult task in EMU design. In this paper, operation principle of four-quadrant rectifier and traction transformer ferromagnetic noise generation mechanism are studied through modeling and simulation in Matlab. Noise design and suppression of EMU traction transformer are described, which provide the references and ideas for subsequent project.

The current mainstream new energy vehicle energy management strategies are mainly designed for a specific working condition without considering the uncertainty of actual driving cycles. Therefore, online identification of driving cycles can lay the foundation for formulating optimal energy management strategies. In this paper, representative driving cycles are selected from the standard library and processed accordingly. Secondly, the K-Means algorithm is used to perform cluster analysis on the processed driving cycles, and two types of comprehensive driving cycle databases were constructed. Then, the driving cycle recognition model based on BP neural network, extreme learning machine (ELM), and generalized regression neural network (GRNN) is constructed in the MATLAB environment to find the driving cycle recognition strategy with high recognition accuracy. Finally, the simulation results show that the operating driving cycle’s recognition model based on the generalized regression neural network (GRNN) can better identify the driving cycles online than the other two types.

As a typical scenario in bus driving, buses entering and leaving stations without speed planning will lead to high energy consumption problems such as frequent start and stop. This research suggests a predictive energy management strategy (EMS) for fuel cell (FC) hybrid electric buses based on speed optimization to lower energy consumption of speed variation while taking into account the features of the vehicle as it enters and leaves bus stops. In the distance of 100 m before and after the bus enters and leaves the bus stop, the speed optimization method based on dynamic programming considers the speed and travel time of the controlled vehicle based on the space step. The results of the simulation demonstrate that the proposed model predictive control (MPC) energy management method, which is based on speed optimization, can maintain the battery's operational status and follow the need for power. At the same time, the proposed strategy can greatly reduce the fluctuation of the FC working range. And compared with the MPC strategy without speed optimization, the energy consumption is reduced by 26.47%.

CaTiO3 perovskite has drawn extensive attention in recent years for its photocatalytic applications related to environmentally friendly fields, such as photodegradation of organic pollutants, carbon dioxide reduction, etc. However, most research was focused on the experimental modification strategies for increasing its photo-catalytic efficiency, little is known about the mechanisms behind it, especially at the molecular level. Our focus here, is the electronic and optical properties of La/Ce-doped CaTiO3 perovskite with different mass fractions. A 2 × 2 × 2 supercell of La/Ce-doped CaTiO3 with concentrations of 2.5 wt% and 5 wt% were modeled and optimized by GGA-PBE functional with a cutoff energy of 571.4 eV in CASTEP. The lattice parameters, electronic structures and optical properties were then calculated. Our results show that the Ce-doped CaTiO3 formed more stable structures than La-doped ones, with the minimum formation energy achieved at a doping concentration of 5 wt%; Additionally, for both La and Ce-doped cases, a higher doping concentration (5 wt%) results in a better potential structure for an n-type material, with its conduction band shifting across the Fermi level. The shift was caused by the adjustment of the band structure with the introduction of rare earth elements, which indirectly influenced the Ti-d-states of CaTiO3. According to our optical property studies, La/Ce doping in CaTiO3 could also reduce polarization and refractive index compared to pure CaTiO3. Our work provides a theoretical framework to study how different types and concentrations of doping elements affect the photophysical properties of CaTiO3 perovskite.

With the continuous promotion and application of new energy vehicles, the demand for charging piles is increasing. In various types of charging piles, the special charging piles of the business circle and private charging piles are idle for a certain period of time, so with the help of block chain technology, a charging pile sharing scheme based on block chain technology is proposed, which creates a convenient, safe and reliable sharing platform on the charging pile sharing system by virtue of the advantages about block chain decentralization, openness, independence, security and anonymity. Under the supervision of the government, users who hold charging piles can apply to join private charging piles to make a certain profit. Besides, we can also make full use of the functions of all types of charging piles, providing more safe and diverse charging schemes for electric vehicles customers. In this way, we can not only give full play to the role of all types of charging piles, but also provide more and more guaranteed services for electric vehicles customers to achieve a mutually beneficial and win-win situation.

To reduce the cost of all-silicon carbide (SiC) MOSFETs based electric vehicle charging station, an electric vehicle charging station based on SiC MOSFETs and silicon (Si) IGBTs hybrid cascaded three-level H-bridge converter is proposed in this paper. The proposed charging station consists of high voltage stage, isolation stage and low voltage stage. In this topology, only two SiC MOSFETs are employed in the high voltage rectifier and all the high frequency switching events are concentrated on SiC MOSFETs by using a specialized modulation scheme. Meanwhile, batteries are integrated in the system which are used as energy storage units. LLC converters are employed in the isolation stage which provide the electrical isolation and power transmission and a low voltage dc bus is built by connecting the outputs of the LLC converter in parallel. The electric vehicles and low voltage energy storages are connected to this low voltage dc bus by using dc/dc converters. A simulation model is built in MATLAB/Simulink and the simulation results show the validations of the proposed electric vehicle charging station.

As the promising device in automotive, stationary, and portable applications, proton exchange membrane fuel cell (PEMFC)’s performance has attracted growing attention, which can be strongly affected by its current distribution. With the characteristic of online and nondestruction, magnetic field can be used to evaluate the current distribution and PEMFC state, while further study is still lacked for understanding the effect of PEMFC stack structure and operation conditions on its magnetic field distribution. In this study, a numerical model of two-cells PEMFC stack is established to investigate its magnetic field distributions at different current densities and states, from which the coupling effect of magnetic field of each cell is clarified. In this PEMFC stack model, the distributions of magnetic field at different states and current densities are constant while the magnitude of magnetic field will vary. Moreover, coupling effect will be weakened with the cell distance between two cells. From the results, these findings can be utilized to monitor the operation state of PEMFC stack in the further researches.

Hydrogen production by proton exchange membrane water electrolyzer (PEMWE) has a good match with wind power and photovoltaics, so in recent years, it has gradually become an important technology for large-scale comprehensive development and utilization of renewable energy. An analysis of the PEMWE modeling method is presented in this paper, and establishes a three-dimensional model of the PEMWE single cell based on the coupling of multiphysics fields, including heat transfer, mass transfer, and electrochemical kinetics. We studied the effects of temperature, pressure, and membrane thickness on PEMWE performance, and determined the optimal configuration of PEMWE single cell based on polarization curves and material distributions in the electrolyzer. Based on the calculation and comparison, under the conditions of 80 ℃ and 40 bar, the PEMWE with 127 μm thick membrane provides better performance, can control costs and meets hydrogen safety standards.

To address the problems of the random and incomplete charging process of the vehicle-mounted lithium-ion batteries, this paper proposes a machine learning method that can realize state of health (SOH) estimation under random charging conditions. Firstly, the complete voltage curve prediction in the constant-current (CC) charging phase under the short-term charging scenario is realized by constructing fitting polynomials, which effectively solves the problem of feature vector acquisition in short-term random charging scenarios. Then, the effects of charging durations of different constant-current charging voltage intervals on SOH estimation are compared to determine the feature vectors. The gaussian process regression (GPR) algorithm is employed to establish the SOH of the battery. Finally, the feasibility of the proposed voltage estimation method is verified at different aging cycles and in random charging scenarios, respectively. The effectiveness of battery SOH estimation based on short-term random charging data is verified. The results show that the proposed method has good feasibility with the SOH estimation error of less than 1.64%.

The classic proportional integral (PI) controller will produce steady-state inaccuracy and poor anti-interference performance while following sinusoidal current commands. On this basis, this paper introduces a quasi proportional resonance (QPR) controller, in which the current inner loop is controlled by a QPR controller and the voltage outer loop is controlled by a PI controller. Firstly, the working principle of bidirectional H4 bridge converter under rectifier condition is analyzed, and the design method of double closed-loop control and its controller is given. The power flow direction of the converter is controlled by voltage regulator, and a set of bidirectional feasible control parameters is derived, that is, the unified control method of bidirectional H4 bridge converter is proposed, The stable control of bidirectional AC/DC in single-phase photovoltaic energy storage system is realized and good dynamic performance is obtained. Simulation and experiments show that the unified control method can realize the seamless switching between rectifier and active inverter.

The problem of optimal siting and capacity allocation of battery energy storage system in power grid is addressed. First, the paper analyzes the advantages and disadvantages of distributed power supply and battery storage system, and then proposes the practicality of battery storage system, second, establishes the mathematical model of battery storage system, lists the optimization conditions and objective function, and introduces the economic cost calculation to determine the optimal site location and capacity configuration by comparing the total cost of accessing different numbers of battery storage systems. Finally, the optimal siting and capacity determination methods of battery storage systems and the characteristics of the calculation cases are summarized and concluded.

As the global energy transition accelerates, renewable energy sources are now widely used in power system. Consequently, accurate forecasting of short-term power load demand and renewable energy sources generation (photovoltaic and wind power) play a key role in energy management system (EMS), power market and grid-building integration. Currently, a large number of separated models about short-term power load demand and renewable energy sources generation forecasting have emerged, while the coupling effect between photovoltaic and wind power hasn’t been considered, which will affect the forecast accuracy. To fill this gap, this paper proposes a physically interpretable coupling forecasting model to explore the coupling relationship, which contains state-space equations and error correction model. The state-space equations investigates the coupling relationship and the error correction model learns the transfer relationship of the forecast error. The numerical results on public datasets show that the proposed method holds great promise, and the proposed forecast model can effectively explore the coupling relationship between short-term power load and renewable energy sources generation, thus can provide accurate and efficient predictions.

A novel active equalization circuit based on ring structure is proposed to solve the problems of over equalization, slow equalization time and inconsistent equalization energy of lithium-ion battery packs. The structure adopts distributed equalization of multiple inductors to quickly realize energy transfer between batteries. The battery state of charge is selected as the equilibrium variable, and the optimization path model is established through the Whale Optimization Algorithm (WOA) to integrate the state of charge. On this basis, the fuzzy control system adjusts the active equalization current according to the input current and input voltage. In order to verify the effectiveness of the proposed topology, the simulation is carried out in the software, and six lithium-ion batteries are built for the equalization experiment. The results show that compared with the traditional structure, it can effectively reduce the inconsistency of the battery pack.

The targets of peaking carbon dioxide emissions and carbon neutrality can be achieved by the large-scale penetration of renewable power production, but the intermittent nature of renewable sources imposes a burden on the operating stability of power system. To improve the peak-shaving capability of power system, a bi-level optimal sizing and dispatch model for hybrid coal-fired power-energy storage system considering different electrochemical energy storage technologies is proposed. The lower-layer scheduling model minimizes the operational cost of thermal power units and penalty cost for unmet load and wind curtailment, while the upper-layer sizing model minimizes the investment cost of energy storage and the overall scheduling cost, which is solved by an iterative method nested with quadratic programming. Finally, the results show that (1) the inclusion of energy storage can eliminate the unmet load and improve power supply reliability; (2) Nickel-Cadmium battery is the most cost-effective option for peak-shaving operation because of its high depth of discharge and long design lifetime; (3) The economic sensitivity analysis of rated power and capacity verifies the optimality of sizing results.

With the adjustment of the energy structure, wind power plants, photovoltaic power plants and other renewable energy power stations are connected to the integrated energy system on a large scale. Based on the typical daily scenario of wind and solar output, a system economic optimization operation method considering market planning is proposed, taking into account factors such as energy storage, grid interaction, energy abandonment penalty, consumption incentive and market planning penalty, etc., to establish a comprehensive energy system economy. According to the optimization results, the output plan of various equipment and the purchase of electricity and gas are reasonably arranged. Finally, through an example analysis based on the actual data in a certain area, it is verified that the optimal dispatching model of the integrated energy system considering the planning penalty can promote the consumption of wind and solar energy and improve the economy of the optimal dispatching of the integrated energy system.

As a new type of energy storage device, supercapacitors (SCs) have the advantages of high power density, long cycle life and wide operating temperature range. However, there is energy loss in the working process of SCs, and the main way is heat loss. Therefore, this paper analyzed the heat generation characteristics of commercial SCs through heat flux measurement experiments, and studied the relationship between the heat generation characteristics of devices and the energy loss of charge and discharge. The results showed that the energy loss increases with the increase of charging current, and the heat loss accounts for more than 85% of the total energy loss. In addition, the total heat generation during the charging and discharging process of the SCs are greater than the Joule heat generation. The difference is partly caused by the irreversible Faraday reaction heat generation and charge redistribution.

In order to deal with the stability and security problems of power system operation brought by large-scale new energy grid connection, this paper proposes a modular multilevel energy storage power conversion system (MMC-ESS) with grid support capability. It utilizes the modular structure of the modular multi-level converter, and connects the battery energy storage in its sub-modules in a distributed manner to form a modular multi-level energy storage power conversion system. By using the access of the energy storage unit, the grid-connected stability of the system can be improved. At the same time, the Virtual Synchronous Generator (VSG) is introduced into the MMC-ESS, so that it has inertia and damping characteristics similar to the synchronous generator during operation, which enhances the power system's ability to deal with frequency disturbances. Simulation results show that the proposed grid-supported MMC-ESS can suppress power fluctuations, provide frequency support, and effectively improve grid stability.

Valve regulated lead-acid (VRLA) battery is in the floating charge state for a long time, and the online accurate assessment of its state of health (SOH) is of great significance. In this paper, the online monitoring platform is built, and the discharge characteristics of battery are tested. Based on the phenomenon of terminal voltage “steep drop and rise again” during discharge, nine characteristics were extracted, including trough voltage, plateau voltage, voltage difference, trough current, plateau current, current difference, trough time, plateau time and time difference. The health factors were obtained by dimension reduction through principal component analysis (PCA) and Pearson correlation coefficient. The BP neural network is built to estimate SOH of the battery and is optimized using genetic algorithm (GA). The accuracy of the battery SOH assessment model is verified by comparing with the capacity check discharge experiment data, and the feasibility of the proposed battery SOH assessment method is also proved.

In the context of high proportion of renewable energy access, in order to promote the consumption of renewable energy, the cooperation between renewable energy stations and energy storage (ES) has become increasingly close. In previous studies, renewable energy stations and ES are often regarded as a whole, but in actual scenarios, the two often belong to different stakeholders. Aiming at the characteristics of distributed autonomous decision-making between renewable energy stations and ES, taking ES allocation capacity and power as shared variables, and decomposing the coordination mechanism based on the alternating direction method of multipliers(ADMM), a distributed collaborative allocation model of two operators is established. Through a small number of iterations of information, the optimal ES allocation strategy under the distributed framework is obtained. Finally, the effectiveness and applicability of the proposed algorithm is verified by comparing the calculation results of the proposed algorithm with the centralized algorithm using the data of a small integrated energy system in North China.

High temperature composite form-stable phase change material is a promising technology to promote and utilize renewable energy and improve the system efficiency. The hygroscopic property of Na2CO3-K2CO3-based composite form-stable phase change material was studied. This study focuses on the effects of material modification and surface treatment on the moisture absorption properties of phase change materials. Experimental results show that the methods of doping zeolite, silica gel and modified magnesium oxide have limited influence on the hygroscopicity of phase change materials. Some methods are not practical because of the limitation of operating temperature. The surface treatment methods can effectively suppress the hygroscopicity of the phase change material. Especially, when the metal film is wrapped, the moisture absorption of the material can be effectively controlled. The combination of coating and wrapping can further protect the material from the moisture as in long-term experiments, the weight gain by moisture absorption is no more than 5%.

Restricted by cost and technology, increasing the power of a single flywheel energy storage device is difficult. Using flywheel array can not only increase the total energy storage capacity of the flywheel system, but also reduce the development and production cost of the unit. For the flywheel array energy storage system, the research on the control strategy of coordinated control and mutual cooperation of each energy storage unit is the solution to realize the efficient and safe operation of the array. This paper firstly discusses the research progress of coordinated control strategies for flywheel array energy storage systems internationally in recent years, and summarizes and analyzes the advantages and disadvantages of various control strategies in flywheel energy storage power distribution and array parallel control. By summarizing and researching the coordinated control strategies of flywheel array energy storage systems in the fields of grid regulation, UPS, rail transit energy recovery, pulse power supply, and integrated energy storage technology, the paper provides reference for the design and innovation of array control strategy of the integrated physical energy storage system.

In recent years, the development of mica capacitor technology has greatly improved the withstand voltage and energy storage density of capacitors, which is suitable for Marx generators. Before using mica paper capacitors to assemble Marx generators, it is important to study the electrical performance and life characteristics of the capacitors. In this paper, a repetitive-rate microsecond pulse test platform was established to research the lifetime characteristic of mica paper capacitors. The test platform is mainly divided into two parts. The first part includes an air-core pulse transformer, a primary electrolytic capacitor, a thyristor and a diode. Another part contains a mica paper capacitor, a SF6/N2 gas switch and a water dummy load. This platform can output voltage up to 60 kV and operate in a repetitive rate of 20 Hz for 50 s or 10 Hz for 100 s. Voltage jitter is lower than 1%. When working continuously, it can charge mica capacitor 2000 pulses for a time. With an interval of about 30 s each time, the test platform can work continuously for one hour.

The safety of lithium-ion battery has received widespread attention. Among them, external short-circuit faults often cause more serious battery safety accidents. In this paper, an electrochemical-thermal model based on Pseudo two-dimensional electrochemical modelling theory and the law of conservation of energy is developed for external short-circuit faults in lithium-ion batteries, and accurate simulation of external short-circuit faults in batteries is achieved through parameter identification. The root mean square errors of the model simulation voltage, current and temperature are 76.9 mV, 46.25 A, and 5.76 ℃, respectively. Based on the established model, the initial SOC, ambient temperature, particle radius of electrode material and exchange current density were selected as the control variables, to investigate the influence of different short circuit initial conditions on the external short circuit characteristics. The results show that the temperature rise rate of the external short circuit of the battery is greater at low initial SOC values and low temperatures.

In recent years, with the development of battery energy storage technology and the support of policy, the construction scale of user-side battery energy storage system is increasing rapidly, and its operation performance has become more and more valued. In-depth quantitative analysis and evaluation is of great significance to provide reliable guarantee for high efficiency, safety and reliability operation of energy storage system. Based on the principles of scientificity, comprehensiveness and operability, the operation evaluation indexes are proposed including charge-discharge performance, energy efficiency, safety, reliability and economic performance. The operation performance of an example battery energy storage system for peak-load shifting is quantitatively analyzed and evaluated, based on the operation data and field test data. And the optimization suggestions are given for the problems existing in the operation of the system. The results show that the proposed operation evaluation indexes and methods can realize the quantitative evaluation of user-side battery energy storage systems on the charge-discharge performance, energy efficiency, safety, reliability and economic performance, which are helpful for the operation and maintenance of user-side battery energy storage systems.

Hydrogen production using renewable energy is an important way to promote new energy power consumption and achieve zero carbon emissions. Compared with the traditional grid-tied water electrolysis, off-grid photovoltaic water electrolysis has the advantages of low cost and flexible deployment. The applicability of three kinds of hydrogen production electrolyzers in combination with renewable energy was investigated. The structure of off-grid hydrogen production system based on alkaline electrolysis water hydrogen production equipment is emphatically expounded. The design methods of photovoltaic DC power supply unit, hydrogen production auxiliary system AC power supply unit and the power supply for control unit of electrolyzer are discussed. The design schemes of capacity allocation of energy storage and the reliable power supply of dual backup are proposed as a technical reference for off-grid photovoltaic hydrogen production systems.

The zero-current opening strategy can effectively improve the electrical life of electromagnetic switches. However, during the period from opening operation to the module sending the opening signal, the zero-current opening strategy requires the control module to have energy storage elements to ensure the reliable maintenance of the electromagnetic switch. The energy storage element increases the volume of the control module, which is difficult to meet the compact design requirements of electromagnetic switches. In this paper, the electrolytic capacitor behind the rectifier bridge in the control module is used as the energy storage capacitor, and a capacity minimization design method based on mathematical model is proposed. By analyzing the charge transfer process of the energy storage capacitor in each working mode of the electromagnetic switch coil drive circuit, building the model of the capacitance and the operation time of the electromagnetic switch after opening operation. According to the maintenance time required for zero-current opening, the minimum capacity of the energy storage capacitor is established to make the volume of the control module as small as possible.

In new energy vehicles, lithium-ion batteries (LIBs) have been broadly applied. Charging anxiety seriously affects the driving experience of passengers, and the battery's charging procedure is intimately associated to the battery's remaining life and safety performance, especially the lithium precipitation of the negative electrode caused by high power fast charging in low temperature environment, which often triggers the thermal runaway of the LIB. The battery equivalent circuit model (ECM) and thermal model are established in this paper, and the battery charging simulation in low temperature environment is carried out in Matlab/Simulink. The optimal frequency is obtained through Mendeley data, and the pulse current and alternating current (AC) with the optimal frequency are loaded. The simulation results show that compared with constant current (CC) charging, pulse current and AC charging have shorter charging time and better temperature rise effect in low temperature environment.

For the contacts of vacuum interrupter with iron core, different parameters of iron core have different effects on the beam magnetic field, in order to study the effect of the optimized iron core structure on the axial magnetic field (AMF) and lag time generated by the contact with the current at peak value and zero. The 3-D model of vacuum interrupter is established, and the iron core structure is compared with different parameters, then the iron core with the best optimization effect is selected for further analysis in this paper. The simulation results showed that when the number of grooves of the iron core is large enough, the effect of iron core fracture to reduce the influence of eddy current on magnetic field is not ideal. After optimizing the structure of the iron core, the AMF generated by the contact is evenly distributed on stationary and moving contact surfaces and on the mid-gap plane, then the magnetic flux density meets the requirements. With the current at zero, the residual magnetic field is small, and the lag time is generally small.

Lithium-ion battery is widely used as a power source in electric vehicles and battery energy storage systems due to its high energy density, long cycle life and low self-discharge rate. Meanwhile, the high inconsistency of lithium-ion battery pack has also attract attention. In this paper, introduce the balanced topology based on various energy storage electronic devices what advantages and disadvantages. The ideas and methods of selecting and improving the balanced topology under different circumstances are compared and analyzed. Finally, combined with the development needs of battery energy storage system in the future, propose the idea of improving the balanced topology.

When DC microgrid is connected to a single-phase AC load, it will cause the DC bus to generate double-frequency voltage pulsation, which seriously affects the normal operation of important units. To solve the above problems, from the perspective of impedance modeling, this paper proposes a secondary ripple coordinated control strategy of DC microgrid based on impedance model. Firstly, the origin of the secondary ripple voltage of the DC microgrid is firstly analyzed, and the input and output impedance models of each unit of the DC microgrid are established. On this basis, the paper introduces band-pass and band stop filters to reshape the output impedance of each key converter at 100 Hz, so as to realize the effective suppression of secondary ripple current by energy storage converter. In order to further reduce the influence of secondary ripple power on each unit connected to the DC bus and ensure the voltage quality of DC bus. The active capacitor converter is introduced to actively control the secondary ripple power by adjusting the virtual capacitance of the active capacitor, so as to suppress the secondary ripple voltage of the DC bus. Finally, the effectiveness of the control strategy proposed in this paper is verified by simulation.

A coordinated control strategy is proposed for the islanded operation of micro-grids with photovoltaic (PV) distributed generation (DG) and energy storage in this paper. Under the premise of exploiting the power generation capacity of photovoltaic distributed generation as much as possible, the control strategy chooses droop control or constant voltage frequency control to maintain the voltage and frequency of the system according to the current operating state. This control strategy can maximize the use of energy exchange ability of every energy storage unit, which ensures the stable operation of the islanded micro-grid system. A photovoltaic-storage micro-grid system model based on this strategy is built on the PSCAD/EMTDC simulation platform to verify the effectiveness of the control strategy.

Smart substations are generally equipped with mechanical characteristics monitoring devices for HV circuit breakers, which collect the key mechanical characteristic parameters of circuit breakers. The key mechanical characteristic parameters of circuit breaker energy storage motor and opening and closing operation are selected for fuzzy comprehensive calculation to evaluate the comprehensive working conditions of the mechanical characteristics of HV circuit breakers. This method takes the error quantity of mechanical characteristic parameters deviating from the normal value as the evaluation factor, determines the membership function according to expert experience, and analyzes and comprehensively evaluates the mechanical characteristics of HV circuit breakers according to the calculation results. In this paper, the on-line monitoring data of the mechanical characteristics of circuit breakers under normal and abnormal conditions are evaluated by fuzzy comprehensive evaluation. The results show that this method can correctly reflect the mechanical characteristics of HV circuit breakers and provides a feasible method for the on-line evaluation of the mechanical characteristics of circuit breakers in smart substations.

Lithium-ion batteries are key components of energy storage systems and electric vehicles, and their accurate State of Charge (SOC) estimation is important for battery energy management, safe operation and extended service life. In this paper, Multi-Kernel Relevance Vector Machine (MKRVM) and Particle Swarm Optimization (PSO) are used to estimate the SOC of Li-ion batteries under different operating conditions. PSO is used to automatically adjust and optimize the weights and kernel parameters of MKRVM to improve estimation accuracy. The proposed method is validated on three battery operation experiment under different operating conditions. The test results show that the proposed PSO-MKRVM can precisely estimate the battery SOC under different operating conditions with an accuracy higher than 0.99 and its maximum average error does not exceed 2%.

Aiming at prominent problems such as thermal runaway triggered by heating, thermal runaway triggered by overcharge and thermal runaway triggered by nail penetration of lithium titanate battery, the paper studied the variation rules of characteristic parameters such as battery surface temperature, environment temperature, battery voltage, characteristic gas and electrolyte vapor and studied the early warning method of thermal runaway of lithium titanate battery. The results showed that both overcharging and heating could trigger thermal runaway of lithium titanate battery, and there was less chance of thermal runaway triggered by nail penetration; a large amount of electrolyte vapor and smoke would be generated in the early stage of thermal runaway; the characteristic parameters of thermal runaway changed dramatically. In the enclosed space, the reference values of the thermal runaway characteristic parameters of lithium titanate battery change like this: The battery surface temperature or the environment temperature can reach 180 ℃, or the average heating rate can reach 5 ℃/s and above and last for 3 s; the battery voltage drops to 0 V in a short time, and the peak rate of decline can reach 0.4 V/s and above; the concentration of H2 can reach 100 ppm, the increase rate of H2 concentration can reach 5 ppm/s and above, and can last for 3 s and above; VOC gas concentration can reach 10000 ppm, and its concentration increase rate attains 1000 ppm/s and above, and can last for 3 s and above.

Accurately simulating the terminal voltage characteristics of lithium-ion batteries in the whole life cycle is the key index to evaluate the performance of batteries. In this paper, Thevenin model is established, and the sensitivity analysis of the OCV and impedance parameters of lithium iron phosphate battery to the accuracy of the model is carried out. Euclidean distance is used to characterize the changes of the parameters of different decay states and new battery models. The results show that with the decline of the battery, the Euclidean distance of thermodynamic and kinetic parameters presents d(RP) > d (OCV) > d (R0); The voltage estimation accuracy of the battery under various working conditions at different stages is verified by the model. The accuracy of the updated OCV, R0, RP, CP model of the battery is 1%, the accuracy of the updated R0, RP, CP model is less than 2%, and the accuracy of the updated OCV model is less than 3%. Therefore, it is recommended to update the R0, RP, CP.

The on-load tap changer (OLTC) is a device which can operate under the excitation or load state of transformer and is used to change the tap connection position of winding. The mechanical fault accounts for nearly 90% of the OLTC fault, so it is necessary to detect the mechanical fault of OLTC. Vibration signals recorded from the operation of OLTCs contain lots of information about mechanical properties. And the mechanical defects can be diagnosed with the help of short-time energy method. In this paper, in order to study the defect simulation and diagnosis method of OLTCs, M-type 220 kV transfer switch and voltage output accelerometer are selected, the tap changer is placed in a 1.98 m high and 1m diameter oil tank. Three typical mechanical defects, such as fracture of energy storage spring, fracture of transition contact guide rod and falling off of fixed contact, are replicated under laboratory conditions. The acoustic fingerprint data of repeated tests of normal condition and 36 tests of defect condition are statistically analyzed. Each vibration signal is processed by the short-time energy method. The vibration characteristics of typical mechanical defects are investigated by analyzing the acoustic fingerprint and short time energy curve.

Internal short circuit is one of the main triggers for thermal runaway of Li-ion batteries, however, internal short circuit is incidental and cannot be predicted in advance. Therefore, improving the sensing capability of internal short circuit is especially important to improve the safety of electric vehicles. In this paper, we propose an algorithm for detecting internal short circuit of Li-ion battery based on loop current detection, which enables timely sensing of internal short circuit of any battery in a multi-series 2-parallel battery module by detecting the loop current. The method only needs to detect the voltage at both ends of the diagnostic resistor (3 measurement points), which has the advantages of fewer detection points and less additional wiring to the battery module, and can avoid the safety hazards caused by excessive wiring while protecting the battery module. The experimental results show that the method is able to detect internal short circuits in parallel lithium-ion battery packs in a timely manner.

With the rapid development of new energy sources, lithium-ion batteries have been widely used in electric vehicles due to their advantages such as large capacity, long cycle life and good temperature performance. Battery health status assessment has become more and more important. The impedance is an important parameter to measure the battery life, its size will change with the use of time, and closely related to the health of the battery. In this paper, the charging curve and Incremental Capacity (IC) curve of lithium batteries are analyzed, and five relevant feature quantities are extracted. By establishing Cuckoo Search Algorithm (CSA) optimized Support Vector Regression (SVR) model to predict ohmic resistance and charge transfer resistance. The CSA-SVR model is verified by experimental data. The mean square error and mean absolute error are less than 6 × 10−6 and 0.2, respectively, indicating high prediction accuracy.

For the sake of measuring the conductivity without damaging the energy storage materials of the energy storage devices, Magnetically mediated thermoacoustic detecting method (MMTDM) is used in this experiment. MMTDM is a new detection technology based on electrical impedance imaging technology and ultrasonic imaging technology, which has the merits of high contrast of electrical impedance imaging technology and high resolution of ultrasonic imaging technology. Its principle is to generate an induced magnetic field by inputting a certain pulsed current to the exciting coil. The energy storage material continuously traps joule heating in the induced magnetic field and expands instantaneously, sending out sound waves reflecting the conductivity information. The sound waves are collected to reconstruct the conductivity information. The paper mainly builds the platform of the MMTDM, compares and analyzes the experimental results of the one-layer excitation system to four-layer excitation system, and observes the strength of the sound waves which are emitted by the energy storage materials under the one-layer excitation system to the four-layer excitation system. This experiment has better verified the feasibility of applying the MMTDM system to test the conductivity of energy storage materials, and provided a relatively perfect theoretical and experimental basis for the development of non-contact testing the conductivity of energy storage materials in the future.

Aiming at the optimization planning problem of mobile energy storage vehicles, a mobile energy storage vehicle planning scheme considering multi-scenario and multi-objective requirements is proposed. The optimization model under the multi-objective requirements of different application scenarios of source, network and load side is established, and the constraints that should be satisfied are given. The optimization solution method based on the second-order cone is adopted, and the large-scale mixed-integer nonlinear model is converted into a mixed-integer second-order cone optimization model, which improves the solution speed of the optimization problem. The calculation example analysis shows that the proposed mobile energy storage vehicle planning scheme utilizes the stored electricity to the greatest extent, and can meet various needs in different application scenarios with limited investment cost.

The pulse power system based on the air-core pulse alternator (ACPA) takes into account the high energy storage density and high power density, has the advantages of miniaturization, lightness and high repetitive output frequency, and is an important technical way of pulse power for electromagnetic railgun. It is necessary to make the speed of electromagnetic rail gun controllable and reduce the muzzle current in the process of actual combat, but the proposed working principle has not yet studied the speed control of rail gun and the strategy of reducing the muzzle current. In this paper, the two-phase discharge process of the ACPA is analyzed through the superconducting loop magnetic chain conservation and the characteristics of its commutation overlap process are analyzed. Then the feedback predictive control of rail gun speed based on pulse wave number control is derived on the basis of two-phase discharge, and the relationship between the initial excitation current of the discharge and the discharge speed of the rail gun is analyzed, and a possible continuous control of the speed of the rail gun based on the pulse power system of the ACPA is proposed.

As distributed energy resources continue to be connected to the grid, the supply side and demand side of the power system are becoming increasingly uncertain in both directions. At the same time, many flexible loads have emerged. We can take advantage of their adjustable characteristics, which can be considered virtual storage to cut peaks and fill valleys for the grid. Data centers and buildings are gradually becoming a hot topic in recent years due to their substantial annual energy consumption. In this paper, we considered a flexible energy aggregator considering virtual energy storage. The cold energy from the data center and the heat energy from the ground source heat pump system(GSHP), i.e., the building, are incorporated as broad energy storage to participate in the aggregator’s dispatch. In this case, the data center(DC) and GSHP hybrid systems have outstanding performance. Then, a two-stage optimal scheduling strategy was used to minimize total cost, which includes day-ahead cost and real-time cost and determine appropriate real-time temperatures for data centers and buildings. We performed a detailed numerical comparison to prove the economy and validity of the proposed model.

The power generation system with renewable energy supply is susceptible to the influence of external environment. Lithium battery and other energy storage devices need to be added in the new energy field to smooth the output of renewable energy generation system and improve the stability of the integrated system. Accurately estimating the state of charge (SOC) of energy storage batteries can effectively improve the use and reasonable scheduling of batteries. This paper takes ternary lithium battery pack as the research object and builds a second-order RC equivalent circuit model to estimate the SOC. According to the time-varying characteristics of battery model parameters, recursive least square method with forgetting factor was used to identify the parameters, the model parameters are modified using the current data. To solve the problem of accumulated error of extended Kalman filter (EKF) algorithm, an adaptive fading factor was introduced to correct prediction error covariance matrix and suppress the influence of historical data on the current state. Matlab simulation and dynamic stress testing experiments show that, compared with EKF algorithm, the adaptive fading EKF (AFEKF) algorithm has higher accuracy.

Valve-controlled battery is the main component of DC power supply system in mainstream substation. The method of determining the inconsistency of battery banks by measuring the capacity of independent charge and discharge experiments has some limitations, and the single battery needs to be separated independently, which is not suitable for the DC battery system running in series online. Aiming at this problem, an inconsistency evaluation model of battery banks for DC power system based on the combination of comprehensive weighting method and grey clustering is constructed. Firstly, entropy weight method and analytic hierarchy process (AHP) were used to obtain the subjective and objective weights of the battery performance parameters, and then the comprehensive weights of the judgment indicators were obtained. Grey clustering was used to comprehensively evaluate the inconsistency of each performance index, and the evaluation model was verified by online detection of battery performance parameters in substation. The evaluation model can evaluate the inconsistency of battery banks under multiple indexes, which provides a practical method and theoretical basis for online screening of backward batteries and ensuring the stable operation of DC battery system.

With the increase of photovoltaic (PV) power penetration level in the system, the requirements for PV integration is becoming stricter to guarantee the secure and stable operation of the grid. PV stations will be possibly required to perform like a synchronous generator which could participate in frequency regulation, reactive power support as well as provide inertia apart from ramp rate control and the energy storage system is a promising solution. A coordinated control strategy for Photovoltaic-Battery Energy Storage System (PV-BESS) based on virtual synchronous generator (VSG) and reactive current injection is proposed in this paper. The PV station is able to provide virtual inertia, deal with energy exchange between PV-BESS system and conventional power grid as well as response to the system frequency change, thus improving the stability of the power system effectively. Moreover, the influences of control parameters on system performance are studied and the optimal configuration method of BESS is illustrated by using sensitivity analysis, providing reference for BESS configuration in the system.

Disorderly charging of large-scale electric vehicles (EVs) will seriously threaten the safe and stable operation of the power grid. This paper proposes a multi-objective optimal scheduling strategy for EVs based on the price-based demand response (PDR) to solve the above problem. Firstly, it combines the period shift model and the user psychological model according to the user trip characteristics of EVs. And the charging load of EVs is obtained by Monte Carlo simulation. Then, the minimum load variance and the maximum customer satisfaction index (CSI) is taken as the objective functions. The non-dominated sorting genetic algorithm (NSGA-II) is used to solve the problem. The optimal compromise solution is chosen as the final solution after the algorithm obtains the Pareto solution set. The results show that the proposed optimization method can balance the relationship between load fluctuation and customer satisfaction.

Lithium-ion batteries are the main power source for electric vehicles (EVs), and the state-of-health (SOH) estimation of lithium-ion battery has become the focus of battery management system (BMS) and has an important role in the field of energy storage. In this paper, a novel deep-learning SOH estimation network is developed, in which convolutional neural network (CNN), bi-directional gated recurrent unit network (Bi-GRU) and squeeze-and-excitation network (SE block) are fused to improve the overall accuracy of the model while improving the ability of the model to reduce the impact of capacity regeneration, and residual connection improves the training speed of the model and avoids the overfitting phenomenon in deep networks. A series of experiments are conducted on two publicly lithium-ion battery datasets, and the experimental results demonstrate that compared with the commonly used deep-learning network-based methods, the proposed method has higher accuracy and robustness.

After the energy storage system is connected to the grid, it can greatly solve the problems of grid loss and voltage fluctuation, but at present, the cost is high and it needs to be optimally allocated, so an optimal allocation method of energy storage based on the sensitivity standard deviation of grid loss is proposed. Firstly, the method uses the sensitivity standard deviation of network loss and Manhattan distance similarity to determine the quantity and location of energy storage access. Secondly, converting multi-objective functions such as node voltage fluctuation, distribution network loss and energy storage capacity into a single objective function using the weighted average method. Finally, a differential particle swarm algorithm is applied to optimize the charging and discharging power of energy storage within 24 h to get the optimal access capacity. The proposed model can greatly reduce voltage fluctuation and the network loss, and have stronger merit search capability by IEEE33 node system simulation.

PMMA is widely used to modulate the energy storage properties of PVDF-based polymers, since it is generally assumed that they present good compatibility. However, recently we found that the compatibility of is PVDF-based polymers and PMMA also influenced by molecular weight. Herein, P(VDF-HFP) composites contained PMMA with distinct molecular weights were systematically investigated, and the impact of the compatibility between P(VDF-HFP) and PMMA on the energy storage properties was revealed. Compared with P(VDF-HFP) composite films with high-molecular-weight PMMA and pristine P(VDF-HFP) dielectric films, P(VDF-HFP) composite films with low-molecular-weight PMMA exhibit enhanced charge-discharge efficiency and resistivity.

Based on the research results of vibration energy collection system and photovoltaic energy collection system, this paper carried out the integration and demonstration verification of vibration and photovoltaic multi-source collection composite micro-energy system. The hybrid micro energy system integrating the solar energy harvesting and vibration energy harvesting is designed and fabricated. The integration and packaging methods of the system are improved. The whole hybrid micro energy system is used as the lumped mass of the vibration energy harvester, which effectively reduces the system volume and mass. Finally, a hybrid micro energy system with a volume of less than 10 cm3 was fabricated, and the ability to supply power to WSN nodes was verified.

Point cloud registration is an important part of 3D point cloud map construction based on multi-line lidar. At present, traditional iterative nearest point algorithm (ICP) is mainly used for point cloud registration. For the traditional iterative nearest point algorithm, point cloud registration is time-consuming and error-prone, which leads to the low accuracy of point cloud registration. Based on the principle of traditional iterative nearest point algorithm, this paper proposes an improved ICP algorithm combining PL-ICP algorithm and PP-ICP algorithm. The map construction is realized by this method, and on the basis of the map construction, NDT algorithm is used to further realize the positioning of unmanned vehicles. The experimental results show that the method can effectively construct the point cloud image, reduce the registration time of point cloud, and has good registration accuracy. Through the validation of the dataset, the positioning experiment can be completed by randomly placing the unmanned vehicle in a certain position on the constructed map.

The continuous operation of railroad engine room equipment is the focus of the railroad industry, which has put forward high requirements for the continuity and reliability of the equipment, and this requirement has driven the research on energy storage units. Energy storage units are mainly realized through lithium batteries, and lithium iron phosphate energy storage systems are expected to be widely used in railroad engine room power storage because of their high energy density and long life. Isolated input-series-output-parallel (ISOP) converters have the advantage of reducing the voltage and current stress of a single power switching device, which can be applied to the lithium iron phosphate energy storage system in the railroad engine room to combine the advantages of both. The normal operation of the ISOP system needs to ensure the input voltage and output current equalization performance of each subsystem, and this paper firstly establishes the small signal model, based on which the voltage and current equalization performance is analyzed, and the transfer function of the system is given, as well as the common duty cycle control strategy suitable for the proposed ISOP system. The simulation results show that the ISOP system has excellent self-averaging pressure and flow performance under the common duty cycle control strategy.

By using adapted YOLOv3 depth network model, an object detection algorithm based on deep learning module is proposed to solve the accuracy problem of insulator detection in aerial images. In order to solve the problems of wrong detection and missing detection of the target, the detection effect of various deep learning algorithms on insulators in complex environments is analyzed, and ResNet-18 is taken as the backbone network structure of the insulator detection model for the power grid patrol inspection. First, the target to be tested is decomposed into multiple component sub targets with intersection, and then it is detected. Then, using the features and meanings of the intersection areas between the components, they are aggregated and redefined, and then a multi-scale feature pyramid is constructed and formed by merging multi-level labels, which is incorporated into the backbone network, effectively using the multi-target multi-level labels to solve problems such as missed detection and error detection, so that the detected target areas are more accurate and meet the real-time requirements. The experimental results show that, compared with R-CNN and Faster R-CNN, the proposed network has better recognition accuracy and detection speed, and significantly improves the accuracy of target detection.

Corrosion detection metal equipment is an important part of intelligent detection of substations. Corrosion detection algorithm based on computer vision and deep learning techniques has achieved preliminary results. The traditional image processing algorithm is limited by the manually designed feature extraction method, which is inferior to the deep learning-based method in generalization performance and robustness. Based on the deep learning methods, the data driven features and strong feature extraction ability make it possible to obtain larger application scenarios. In this paper, a two-color space-based corrosion defect detection method is proposed to recombine the effective features in different color spaces to help detect rust defects. We used two lightweight models to extract features of each color space and then aggregated the features of the two-color spaces based on channel exchange. Finally, the location and classification of rust defects were completed on the aggregated features. In this paper, the effectiveness of the proposed method is verified on real datasets, and the comparison with the single-color space method proves that the proposed method has better performance.

Pin is an important fixed component of high voltage line. Therefore, pin defect is an important component in the safety inspection of high voltage line. Due to the long line and harsh environment, it is difficult to obtain a large-scale dataset of pin defect. The detection with transformer (DETR) based method has achieved great success on large-size datasets, but it is less effective on small-size datasets. In this paper, we found that one of the main reasons for the poor performance of DETR-like method on small-size datasets is that the one-to-one matching of Hungarian matching leads to the small number of positive examples during training, which makes the model difficult to converge. To solve the above problems, we propose data-efficient Hungarian match (DEHM) and group object query (GOQ) to increase the number of positive examples in training. DEHM and GOQ will not add any parameters during training, and will not affect the inference speed. Extensive experiments show that DEHM and GOQ can improve the performance of DETR-like methods on both small-size dataset to achieve similar results to those on large-size datasets.

The substation’s stability and reliability are of great importance in the power transmission line. The equipment in the substation has high requirements for insulation performance and mechanical properties, so there are many equipment failures in the substation. It is necessary to find faults in time, confirm fault types, analyze fault causes and eliminate the impact of faults. In this paper, we propose a heterogeneous parallel computing based thermal fault detection method using infrared and visible image fusion. Our method is deployed on heterogenous computing platform using CPU and FPGA, which can improve the energy efficiency. We connect multiple FPGA cores and optimize the CPU task scheduling mechanism to achieve low latency and high computing efficiency. Meanwhile, our method adopts deep learning models to extract and fuse the features of infrared and visible images. By combining multi-modality features, the thermal faults detection is conducted fast and accurate. The final output of our method is the fused image with abundant information and the thermal faults are bounded with boxes. Since our method adopts popular deep learning framework, the deployment on CPU and FPGA platform is effecient. Experiments demonstrate that our methods could effectively detect the thermal faults of power equipment of substation and show the superiority than state-of-arts methods. Especially, our heterogenous CPU and FPGA computing platform has huge advantage on energy efficiency.

The application of image recognition technology to weapon systems has not yet formed a systematic theory. In order to study the high-precision identification of projectiles in weapon systems, the application of projectile identification in engineering is realized. In this paper, the recognition technology based on deep learning is carried out, and on the basis of the traditional algorithm, the characteristics of the appearance image of the ammunition are analyzed, and combined with the complexity of the appearance detection of the ammunition, an image processing technology of undercurrent channel filtering combined with YOLO V3 is proposed, and the expected results are obtained after verification based on video sequences and images. Experiments show that the algorithm has relatively high applicability, more stable detection effect and higher accuracy.

This paper focuses on the functional safety of the electric vehicle drive system. Taking the permanent magnet synchronous motor (PMSM) as the research object, the research on the three-phase current response and suppression of the PMSM in the active short circuit (ASC) mode is carried out. In this paper, a transient analytical model of PMSM is established, and the response of three-phase currents after PMSM enters ASC mode is deduced and analyzed. Then, the influence of the phase angle ( $$\delta$$ δ )of the PMSM at the moment of ASC on the current peak value Iabc_max is discussed. The real motor controller and motor simulator are used to test and verify, and the results show that $$\delta$$ δ does affect the size of Iabc_max. In order to reduce the impact of excessive current on other devices, the value of Iabc_max should be as small as possible. The mathematical model of the three-phase current after ASC is established in MATLAB, and the $$\delta$$ δ corresponding to the minimum value of Iabc_max is obtained, so it is only necessary to control the motor to enter the ASC at the moment of $$\delta$$ δ , the value of Iabc_max can be reduced, and the safety and reliability of the motor controller can be improved.

Proton exchange membrane fuel cells (PEMFC) are widely used in various fields due to their low operating temperature, high reliability, fast start-up, and long life. However, fuel cell technology is still severely constrained by its electrical efficiency, price, and lifetime. This paper proposes a power distribution method for multi-stack fuel cell (MFC) system to optimize efficiency and lifetime. A high-efficiency working range is defined to ensure the system efficiency. The startup sequence of each stack is determined by the voltage degradation rate to achieve aging consistency. Three typical methods are compared with the proposed method by simulation, and results show that this distribution method increases the lifespan by 81–87% and the efficiency by 3–3.5%.

Health monitoring of lithium-ion batteries is a major task to ensure the performance and reliability of electronic vehicles. A precise state of health prediction is still a challenging problem caused by the distribution discrepancy between training and testing. In this paper, a domain adaptation network is developed to address the domain shift problem in battery SOH estimation. First, the CNN-GRU network is used to construct the complex mapping from measured samples to capacity. A hybrid domain adaptation method, combined with feature alignment and domain-adversarial leaning, is then used to reduce the distribution difference between domains. The superiority of the designed approach is demonstrated on three NASA battery datasets. Comparison tests indicate that the developed method can obtain precise and robust estimates of the SOH prediction for different cells within the limited target battery cycle data.

The cascade utilization of retired lithium batteries to build an energy storage system is an effective means to achieve my country's dual-carbon goal, but safety issues restrict large-scale promotion and application. Accurately assessing the operational risk of cascade batteries in an energy storage system can ensure the safe operation of the system. This paper defines the risk of retired power batteries in the energy storage system, and establishes the risk with the remaining useful life (RUL), state of charge (SOC)and temperature rise rate of the echelon battery as the evaluation factors. Evaluate the model. In this paper, the BP (back propagation) neural network algorithm is used to estimate the RUL of the echelon battery, and the nonlinear model of the echelon battery is used to estimate the SOC and the temperature rise rate. Combined with the AHP (analytic hierarchy process) method and K-means mean aggregation. The class method estimates the subjective and objective weights of the evaluation indicators. This method can complete the risk assessment and determine the warning threshold value, and finally realize the real-time operation risk estimation during the operation of the echelon battery. The calculation example shows that the method can realize the operation risk assessment of the cascade battery energy storage system, improve the safety of the system, and promote the large-scale popularization and application of the cascade battery energy storage system.

To analyze the impact of power flow transfer of the interline power flow controller (IPFC) on safe operation, this paper studies the safety boundary constraints of the IPFC converter station based on the PQ power domain, the impact on the safe operation of IPFC grid when power flow transferring is analyzed. Nevertheless, to solve instability caused by the power flow transfer, a virtual impedance-based method is introduced to increase the safety boundary and enhances the stable operation ability of the system. Finally, the simulations of a 3-terminal IPFC system established in PSCAD/EMTDC software verify the correctness of the proposed method.

Thermal pressures in the cable’s long-term operation play an essential role in the insulation performance of the high voltage cables. A seven-year serviced and a new 110 kV XLPE cables were selected in this paper. Firstly, they were thermally aged at 110 $$^\circ{\rm C} $$ ∘ C for 60, 120, and 180 days by an operation simulation aging test. Then, the outer sheath was removed, and the cable was peeled and annealed at a range of 90 to 115 $$^\circ{\rm C} $$ ∘ C with a gap of 5 $$^\circ{\rm C} ,$$ ∘ C , respectively. The differential scanning calorimetry, gel content, and DC conductivity were measured. The results show that 60 and 120-day-aged samples are beneficial for crystallization and show lower conductivity; 180-day-aged samples show a slight difference in crystal characteristics from fresh samples and give higher conductivity. After thermally annealed, 90–105 $$^\circ{\rm C} $$ ∘ C always promotes better crystal characteristics and lower conductivity changes. For the new cable, the optimum annealing temperature corresponds to the highest melting point and crystallinity, and the lowest conductivity is 105, 105, 100, and 95 $$^\circ{\rm C} $$ ∘ C , related to the increasing aging days. The same results appeared in the seven-year serviced cable, but the optimum temperature was reduced to 95 $$^\circ{\rm C} $$ ∘ C for both 120 and 180-day-aged samples. In summary, the crystal characters and the thermal-electrical performance of XLPE insulation from the two selected cables could be improved if they were annealed at proper temperatures regardless of the various degradation.

Variable speed pumping and storage(VSPS) can control active power indirectly by controlling rotor speed through converter under pumping condition. In the current modeling of VSPS under pumping condition, most of the optimized formulas of guide vane opening and speed adopt linear approximation, which leads to the fact that the actual absorbed active power cannot fully respond to the command value. Therefore, this paper presents a modified method of guide vane opening and speed optimization. Firstly, the simplified electromechanical transient model of VSPS under pumping condition is established. Then, the operating range and characteristics of the pump are obtained by simulation experiment. The linear expressions of guide vane opening, rotational speed and active power are obtained by linear fitting. The active power in the fitting process was modified from the actual value to the command value to obtain the cubic polynomial. The simulation results show that the modified active power can track the instruction value well, and the difference between the active power and the command value is an order of magnitude smaller than that before the modification.

Aiming at the problems of short driving range and insufficient power of electric vehicles, this paper proposes an energy management strategy of hybrid energy storage system based on multi-objective model predictive control. Firstly, on the basis of the semi-active hybrid energy storage system topology, a Lithium-ion battery model and a DC-DC converter efficiency model are established. Then, aiming at reducing the power loss of the energy storage system and stabilizing the bus voltage, a multi-objective evaluation function is constructed, and the constraints of the control variables are considered as the constraints, and the particle swarm optimization method is utilized to obtain the optimal solution. Finally, experiments are carried out under HWFET driving cycle. The experimental results show that the proposed strategy can achieve optimal power distribution while maintaining the stability of the bus voltage, and make the system efficiency reach 91.79%. It effectively avoids the impact of large current on the life degradation of Lithium-ion battery, and reduces the power loss of the hybrid energy storage system.

The goal of this study is to develop a coordinated cathode and anode pressure management strategy for proton exchange membrane fuel cell (PEMFC), to begin with, a PEMFC system model is created, and adaptive fuzzy PID control and PID control strategies are proposed. This approach controls the difference in pressure between the cathode and the anode. Simulations are used to compare and assess PID control and adaptive fuzzy PID, according to the analysis's findings, the fuzzy adaptive PID technique responds more quickly and experiences less overshoot when loads change. Anode pressure following cathode pressure can prolong the PEM fuel cell's working life and maintain it reliable operation.

The battery packs experience alternate current in the modular multilevel converter battery energy storage system (MMC-BESS), which can cause additional charge throughput and shorten the lifetime of the battery. Therefore, an additional charge throughput reduction method has been proposed for the MMC-BESS based on the second-order circulating current injection. Firstly, the principle of the MMC-BESS is introduced. Then, the additional charge throughput is analyzed by the instantaneous power fluctuation in the arms of the MMC-BESS. Afterwards, a second-order circulating current is injected to adjust the power in the arms of the MMC-BESS. As a result, the additional charge and discharge of the battery packs can be reduced in each fundamental cycle. Finally, comparative experiments are conducted and the results show that compared with the conventional methods, the proposed method can effectively reduce the additional charge throughput of the battery packs in the MMC-BESS.

The longer measurement time and high equipment cost of traditional electrochemical impedance spectroscopy (EIS) testing techniques make it difficult to be applied to engineering practice. To address the above issues, firstly, this paper proposes a rapid impedance spectroscopy testing method based on M sequence. Three sets of M sequences with different frequencies and orders are applied to the battery in turn as current excitation signals to get the corresponding voltage response signals, and then the impedance information of batteries at each frequency is obtained using the continuous wavelet transform algorithm, which greatly reduces test time. Secondly, according to the characteristics of the impedance spectroscopy of LiCoO2 batteries, a third-order fractional-order equivalent circuit model is established, and the impedance spectroscopy of each aging state is fitted to obtain the corresponding model parameters. After the correlation analysis of parameters, this paper builds a multiple linear regression equation between internal resistance parameters and state of health (SOH) for SOH estimation. Finally, it has been verified that the error of the proposed SOH estimation algorithm is within 0.3%.

With the introduction of high proportion of renewable energy and power electronic equipment, more and more complex multi-band oscillation accidents occur in power systems. Multi-band oscillations may damage power equipment and endanger the stable operation of subsystems or even the whole power system. To solve the above problems, this paper proposes a two-stage multi-band oscillation online monitoring method combining synchronous wavelet compression transform (SWT) and Gaussian Naive Bayes (NB) algorithm. The first-stage classifier determines the oscillation type and generates an alarm. If the sample is judged to be low-frequency oscillation, the sample is input to the second-stage classifier to identify the attenuation coefficient and damping ratio, and determine the specific oscillation type. The proposed method can solve the problem of unbalanced oscillation samples, and avoid the problem of manually setting alarm threshold, which is beneficial to improve the accuracy of identification. The simulation results of the four-machine two-area system with doubly fed induction generator (DFIG) show that the proposed method can take into account both the rapidity of oscillation alarm and the accuracy of oscillation identification, and has the function of precise parameter extraction for low frequency oscillation signals.

For the upkeep of electric vehicles, the real-time evaluation of the state of health (SOH) for lithium-ion battery is crucial. This paper initially investigated the viability of estimating SOH based on charge curves because the battery of an electric car has instability, large load, and inconvenient operation while measuring the discharge capacity. The SOH model based on charge capacity was then developed, and it relates the battery charge capacity to the time needed to fully charge the battery. In order to solve the issue of estimating the SOH of fragment charging data, an estimation algorithm based on the fusion of convolution neural network and long short-term memory neural network was presented. The proposed estimation model with deep learning is shown to be efficient and real-time in the SOH estimation of lithium-ion battery using the two battery datasets.

Numerical simulation method has been conducted in this paper to investigate the cooling and heating performance of liquid cooling adopted in Lithium-ion battery pack under typical cooling operating conditions of high-speed climbing, overspeed and driving durability for an electrical vehicle. In addition, the numerical simulation results have been validated by experimental data of battery temperature rise under high-speed climbing operational condition. The results indicate that simulation results are in good agreement with experimental data, and the maximum prediction deviation is 5% compared with experimental data. Numerical simulation model can effectively predict the temperature rise characteristics of lithium-ion battery. The temperature distribution characteristics of battery cooling plate, lithium-ion battery pack and the middle plane section of battery cells seem to be similar at high temperature cooling operational conditions, which is determined by lithium-ion battery pack cooling system structure. The heating temperature rise rate of lithium-ion battery pack can reach 0.95 ℃/min, and the maximum temperature difference of the battery pack during heating process is 10.4 ℃. The low efficiency heating region outside of the battery pack can be improved by increasing the corresponding area of the battery cooling plate and increasing the thermal protection of the battery cells.

With the development of technology in the past decade, China has gradually occupied an absolute leading position in the field of electric commercial vehicles, and has achieved great success in electrified buses. However, the mature powertrain of electric buses, including the single motor direct drive and single motor equipped with reduction gear, cannot meet the needs of electric duty trucks. Though automated mechanical transmission (AMT) is a feasible solution, the power interruption and shift shock hinder its further application. Also, using dual-motor is a promising way to reduce the motor size of duty trucks. It is challenging to find an applicable and optimal architecture for duty trucks due to the diverse potential topology. In this paper, a novel dual-motor coupling propulsion system is proposed, to take advantage of lower mechanical gear pairs reaching more significant driving modes, a special composite transmission is developed. The basic configuration and operating principle are illustrated. As for the particular transmission, a dedicated gear ratio design and optimization process is put forward. Compared with traditional powertrain configuration, the proposed propulsion system can are quite compact, but can realize satisfied dynamic performance and better energy-saving potential.

With the rapid development of shipbuilding technology, the number of large tonnage ships at home and abroad has increased dramatically. In order to reduce the pollution caused by diesel generators in the port industry, promoting the use of shore power by ships is one of the key tasks of energy conservation and emission reduction in major ports. However, the uncertainty caused by waves and other factors in the operation at sea makes the stability of the interconnection of ship-shore cables decrease. While considering how to improve the stability of shore connection, we should also explore how to maintain the power supply of ship power system when the shore connection is disconnected. Therefore, this paper uses Matlab/Simulink simulation software to simulate and analyze the marine electrochemical energy storage system under ship-shore connected cable faults. Once the cable connection failure, diesel generator or battery energy storage system could be used as a standby power supply to serve as the emergency power supply. By analyzing the advantages and disadvantages of the two standby power sources, the priority application status of the battery energy storage system in the fault scenario of ship-shore connection is first established because of the fast response ability of electrochemical energy storage system. Secondly, compared with diesel generators, the simulation results proved that the energy storage system can not only quickly make up for the power supply gap, but also save electricity costs, improve resource utilization, and reduce noise and environmental pollution when the cable fault occurs in the ship.

In this paper, a new method of sodium-ion battery SoC prediction based on recurrent deep forest is proposed. The method uses data that is easy to be measured online, such as voltage, current, voltage and current at the previous moment, as the input characteristics of the model. The predicted value of the SoC is also used for the input of the next moment. The model used a few of experimental data of sodium-ion batteries for offline training, then used the rest of battery data for online prediction. We set up two sets of experiments to test the accuracy of the model from different periods of the sodium-ion battery cycling. The results proved the reliability of the recurrent deep forest method. The experiment also proves that the method has good generalization ability, which surpasses technologies such as Gaussian process regression, reaching the level of advanced machine learning techniques. The details of results indicate an opportunity for future optimization of the algorithm proposed.

The safety of lithium-ion battery has attracted much attention in recent years. Early warning of battery failure is helpful to the safe operation of the battery system. In this paper, the transmission characteristics and failure mode of lithium-ion batteries are analyzed. A simplified finite element simulation model of battery ultrasonic transmission characteristics is constructed. By comparing and analyzing the ultrasonic transmission characteristics of normal battery cells, batteries with inconsistent structures, and batteries with defective structures, the relationships between the inconsistent conditions, fault conditions and feature parameters of batteries are given. The research provides support for battery fault diagnosis technologies based on the ultrasonic nondestructive testing methods.

On the premise of meeting the dynamic performance requirements of all-terrain vehicle, the multi-objective optimization function is established with the energy, volume and cost of lithium battery pack as evaluation indexes. Firstly, according to the vehicle dynamic equation, the basic parameters and dynamic performance indexes of the vehicle are given to calculate the required power and energy of the vehicle. Then, the constraint equations for the voltage and volume indexes are obtained based on the technical requirements of the battery system. Finally, a parameter matching method of lithium battery pack based on multi-objective optimization algorithm is proposed. The results show that the proposed parameter matching optimization method for ATV battery system is reasonable and effective.

Membrane water and liquid water of Proton exchange membrane fuel cell (PEMFC) have hysteresis, overshoot and uneven distribution under dynamic conditions. These characteristics lead to dehydration or water flooding which affect the performance and working life of PEMFC. In this paper, a three-dimensional non-isothermal two-phase model of PEMFC was established, considering the transport of water content and the liquid water saturation. The model was simulated under 0.8 V → 0.5 V → 0.8 V. The results showed: At 0.8 V → 0.5 V, the lag time of average water content in ACL was 25 s, and the location of the lowest water content was inlet in centerline between ACL and MEM. At 0.5 V → 0.8 V, the lag time and the overshoot of average water saturation in CCL were 45 s and 9.5%, and the location of the highest water saturation moved from inlet to outlet at the edge of interface between CCL and MEM.

Air express transportation is an important parts of air transportation, and most of air express are loaded in air containers. At present, air express relies on staff to complete the operation with experience and operating habits when loaded in air container, which restricts the transshipment efficiency of air express transportation, and it is urgent to propose an intelligent loading method of air container. Based on the demand of air express transportation, this research is carried out on the air containers loading problem for express package. Fully considering the characteristics of irregular air container, an intelligent loading method of air container based on modular assembling is proposed to improve the efficiency of loading decision through modularization and standardization. Experimental results show that the proposed method can effectively improve the loading level of air container, and the average volume utilization rate is more than 90%.

To settle the problems related to passenger safety in rail transit stations, improve the management of areas lack of security service, and implement the intelligent detection of passenger flow safety status in subway stations, we have established a passenger safety status detection system for rail transit stations based on machine learning for target detection, passenger safety status detection, terminal data visualization. Human pose estimation based on YOLOv5 and Alpha Pose enables intelligent detection and optimization of the adjustment model through training of passenger behavior data sets. A non-open terminal visualization platform is established to detect the safety status in each monitored area in real time, and provide real-time safety status information and corresponding areas for station staff passengers, so that managers can timely confirm and process passengers with abnormal behaviors. The abnormal behavior data of passengers in rail transit stations are statistically analyzed to evaluate the safety status of passenger flow in stations and provide data support for improving safety management level. The research results will help to improve the intelligent level of passenger safety management in rail transit stations, which is of great significance to ensure the safe operation of urban rail transit.

Railway container transportation has gradually become an important direction for the development of railway freight transportation due to its advantages of simplifying operation and saving cost. This paper mainly studies the acceleration and reinforcement force of the freight in the container when the container is transported by railway flat car. There are specific force calculation standards for the loading of freight on railway vehicles in our country, but on the loading and reinforcement of freight in containers, the relevant railway rules and standards in our country are vague and lack of quantitative provisions. In addition, due to the problems of fit and tolerance in design, there will be a clearance between the corner fittings of the container and the locks of the flat car after the container is loaded onto the car. This clearance will be eliminated first when the car is impacted. This process will play a certain buffering role in the loading and reinforcement of the freight in the container. This paper uses SIMPACK to create an impact model, analyses the impact of the clearance on the longitudinal acceleration of freight with different impact speeds, providing reasonable suggestions for railway container transportation safety in theory.

With the development of economic globalization, more domestic goods need to be transported abroad, but the southwest inland region of China is unable to transport goods from this region in time. In this case, China put forward the overall plan for the New Western Land-Sea Corridor in 2019, aiming to strengthen the development of the West. This paper aims to complete the transportation of container goods under the requirements of low cost and fast time, and establishes a container multimodal transportation model based on the time value of cargo. The specific contents include: firstly, introducing the current development status of the New Western Land-Sea Corridor; Then, the factors that affect the optimal route selection are analyzed; Then, establishing the general model of container multimodal transport based on the time value of goods and solving it by using the idea of ant colony algorithm; Finally, taking the shipment from Xi’an to Singapore as an example, finding the optimal route is through the New Western Land-Sea Corridor, and analyzing the results.

The purpose of this research is to obtain piezoelectric actuators and their application in active vibration control systems, as well as relevant theoretical and practical knowledge, and to use the controllability Gramian method to perform piezoelectric actuators on cylindrical structures through Gramian matrix. The optimal position is determined, and then the finite element analysis module of ANSYS software is used to simulate, and the actual phenomenon is deduced by analyzing the simulation results, and the simulation analysis of the active vibration control of the cylindrical structure is carried out. Then use the relevant knowledge to determine the controllability and observability of the system, and finally use MATLAB for calculation, and perform the final simulation control according to the calculation to complete the control of the vibration of the cylindrical structure and the determination of the optimal position of the piezoelectric patch.

In recent years, the construction of China’s high-speed railway network has been fruitful and the express market has developed rapidly. China Railway Express (CRE) has launched various high-speed railway express products a few years ago, but market share of these products has not been high and problem of freight collection and distribution is not solved well. On the basis of existing network, high-speed rail express nodes are divided into three levels and five indicators in this paper. The process of collecting and distributing transportation is transformed into location-allocation problem, and based on this, a mixed integer programming model aiming at minimum total cost is established on the premise of meeting the needs, and the Lingo software is used to solve this problem.

With the development of the air transport industry, the regional multi-airport or dual-airport system have gradually developed in China, thus becoming a new feature of China’s civil aviation development. Multi-airport system is able to enlarge the air transport hinterland of the city and provide more alternatives of air transfer services. The flight connections between the dual- or multi-airport systems shall be analyzed and evaluated to judge the current operation efficiency and guide the airlines to coordinate their service schedules between the airports. This paper adopts a connection quality index calculation method to evaluate the flight connection quality of Beijing dual-hub system, i.e., the Beijing Capital Airport and Beijing Daxing Airport, and that of different operating airlines at these two airports.

With the rapid development of high-speed railway, its operation management and traffic safety are becoming more and more important. Foreign matter intrusion around high-speed railway needs to be timely prevented and identified. The current intrusion target recognition system still has the problems of high miss rate and false detection rate for the recognition of small intrusion targets beyond 100m. In this paper, aiming at this practical problem, combined with the deep learning target detection algorithm and the requirements of high-speed railway perimeter prevention and control, a system with better intrusion target recognition effect is designed to make up for the shortcomings of current high-speed railway perimeter video monitoring. Based on the actual video monitoring picture of high-speed railway, this experiment establishes a data set of far and small intrusion target recognition of high-speed railway perimeter, and establishes a complete evaluation standard system. Through the selection of algorithm model, it improves the ability of far and small intrusion target recognition of high-speed railway perimeter, and combines the algorithm to establish a far and small intrusion target recognition system of high-speed railway perimeter.

In this paper, an expression recognition algorithm based on multi-source image fusion detection is designed. The visible facial expression image and thermal facial expression image obtained by multi-sensor are used to train the expression recognition model of two channels, and the yolov5 target recognition algorithm is used to identify and classify the photos of the two modes respectively. In the part of modal fusion, decision-level fusion is used to fuse the classification results of the two channels to integrate the respective advantages of the two modes. The experimental results show that the fusion algorithm can have higher recognition accuracy and reduce the occurrence of misjudgment.

With the development of the economy, the demand for township logistics has greatly increased. As the end link of the transportation process, the efficiency of logistics distribution directly affects the transportation cost and customer satisfaction. Compared with urban logistics, township logistics has the characteristics of fewer distribution centers, more decentralized logistics demands, and longer distribution distances, making it difficult to achieve point-to-point distribution. Meanwhile, there is less traffic congestion on the way of township logistics distribution, and the linear relationship between distribution distance and distribution time is more significant, so shortening the distribution time can be simplified as optimizing the distribution path. This paper collects the data of the distribution center in a township and the villages under its service, and uses the simulated annealing algorithm to optimize the distribution path. The optimization results show that the optimized path is 4.3% shorter than the original path, and the optimization effect is obvious.

Assembly pressure is one of the main factors affecting the internal charge, heat, and mass transfer of proton exchange membrane (PEM) fuel cells. It causes the deformation of the components to change the electrical and gas transport characteristics of the internal contact interface of the fuel cell, and affects the overall performance of the cell. In this paper, a three-dimensional finite element model of a single cell considering the surface topography of the component and the contact behavior of the asperities is firstly established, and the contact pressure and diffusion layer porosity of the contact interface of each component of the fuel cell are obtained; Then, the total contact resistance of the cell and the effective diffusivity of the gas in the diffusion layer were established; Finally, a detailed voltage model considering different losses of the fuel cell was established based on the MATLAB platform. The results show that the contact resistance decreases with the increase of the assembly pressure, and the maximum error between the calculated contact resistance and the experimental value is 1.7%, which meets the accuracy requirements; when the clamping pressure is 0.75MPa, the sum of the ohm internal resistance and the concentration difference internal resistance is the smallest, the fuel cell output power is the largest, which can provide guidance for the assembly of fuel cells in practical applications.

In order to further improve the information analysis and mining capabilities as well as the data sharing and interaction capabilities of the smart grid, a non-intrusive load monitoring system with complete functions and strong practicability is designed. Based on the analysis of system requirements, complete the design of the system technical architecture and overall system plan. Start with the data acquisition module design, main control module design, communication module design and other aspects to complete the design of the core functiongal modules of the system. Finally, test the operating performance of the system and the online monitoring function of the clound platform. The results show that the non-intrusive load monitoring system designed in this paper runs normally, reliably and stably, and the realization of each functional module meets the relevant design requirements.

The concept of digital twin provides a new solution for the condition monitoring of DC bus capacitor bank. In this paper, a condition monitoring method of DC bus capacitor bank based on digital twin model is established. Firstly, based on the analysis of the physical characteristics of the capacitor, the digital twin model framework of the DC bus capacitor bank is constructed. After that, the application details of the proposed digital twin model in the condition monitoring of the DC bus capacitor bank are analyzed from five steps, which include digital modeling, synchronous operation, accuracy judgment, parameter calibration and condition monitoring application. Finally, the proposed method is verified by simulation.

With the development of science and technology, energy storage capacitors are gradually developing towards miniaturization and high temperature. Recently, polyvinylidene fluoride (PVDF) has attracted wide attention due to its high dielectric contant and high energy storage density. However, the problems of high dielectric loss and low temperature stability restrict the application of polyvinylidene fluoride. In this paper, it is proposed to optimize the dielectric loss and breakdown strength of polyvinylidene fluoride-based energy storage materials by blending polymethyl methacrylate (PMMA) with high glass transition temperature and high breakdown strength performance. On this basis, the crosslinking agent 1,6-hexanediamine was used to realize the crosslinking of PVDF and PMMA to improve the temperature stability of the material. Through this scheme, the finally obtained crosslinked PVDF/PMMA (40/60) film has an energy storage density of 10.4–11.9 J/cm3 at 30–90 ℃, and efficiency of 79–88%, which are better than most dielectric polymers. Our work provides a solution for optimizing the temperature stability of the energy storage properties of polymer ferroelectric materials to achieve higher energy storage densities at higher temperatures.

At present, pure electric vehicle technology is still in an immature developmental stage. As a transitional technology, hybrid vehicle technology is still of great research significance. In this research, the configuration and mode-switching of a dual-clutch transmission (DCT) hybrid electric vehicle were deeply studied, and a pre- and post-switchable configuration with better comprehensive performance was proposed. A traditional DCT fuel vehicle model and pre- and post-switchable configuration vehicle model were built using the CRUISE software. MATLAB’s SIMULINK software was used to design a rule-based mode-switching control strategy and complete a joint simulation with CRUISE. Using the ISIGHT software, the multi-island genetic algorithm was used to develop a mathematical model to optimize the parameters of the transmission system, and a joint simulation was conducted again. Two simulation results show that the pre- and post-switchable configuration could significantly improve vehicle economy compared with the traditional fuel vehicle configuration.

Lane detection is the core problem of autonomous driving. After completing lane recognition, the autonomous driving system can realize the active safety and control function of vehicle lateral movement. However, the existing methods cannot adapt well to various environments and generate many unnecessary points, resulting in low detection accuracy. In this paper, Point Instance Network (PINet) based on key points estimation and instance segmentation is used, which is composed of several stacked hourglass networks that are trained at the same time. Compared with existing algorithms, PINet achieves ideal accuracy and false positive rate on CULane, especially in night and dazzle light.

Li-rich Layered Oxide Cathodes materials (LLOs) are attractive to researcher by reason of their high specific capacity and low cost. But LLOs has poor rate performance and cycling performance due to its poor dynamic performance and low conductivity. In order to improve the rate performance and cycling performance of LLOs, the lithium rich materials were synthesized by coprecipitation method and solid-phase method. Then coating carbon nanotube graphene hybrid on the surface of LLOs, and the alternating vertical multi-level conductive paths were realized on the surface of the materials. The average capacity at 5C high magnification was 142 mAh/g, which is better than that of the uncoated one of 117.3 mAh/g; Accordingly, the capacity ratio of high and low magnification is increased from 48.85% to 54.6%. The multi-stage conductive path constructed by coated carbon nanotube graphene hybrid accelerates the electrochemical reaction kinetics of the electrode, which is beneficial to improving the ion transmission of lithium ions and electrons. This will effectively improve the electrochemical performance of LLOs.

Accurate prediction of the capacity of lithium battery is of great significance for the failure prediction and health management. At present, the prediction accuracy of machine learning cannot meet the demand, so this paper proposes a prediction model based on genetic algorithm (GA) to modify extreme learning machine (ELM). Firstly, the characteristic quantities related to degradation of battery are extracted from the cycle life test; then, the aging characteristic parameters of the battery are analyzed using the correlation coefficients; finally, a GA-ELM neural network model is established and the input weights and implicit layer bias of the extreme learning machine are optimized using the genetic algorithm. The proposed model is validated with the experimentally obtained battery data and compared with a single ELM neural network model prediction method. The results show that: the method proposed in this paper can effectively predict the state of health (SOH) with better prediction accuracy than the single ELM model.

For long-distance linear motion fields, such as rail transit, electromagnetic ejection, material transmission and automatic industrial manufacturing line, PMLSM is irreplaceable due to the superiority of direct drive. In this paper, in order to suppress the thrust fluctuation of discontinuous segmented linear motor (DSPMLSM), the thrust fluctuation model of DSPMLSM is obtained. Then, an experimental platform is built to verify the theory analysis, the thrust fluctuation data is collected at low speed. In order to analyze the thrust fluctuation appropriately, the regional fast Fourier transform (FFT) is proposed which can effectively reduce the harmonic number of thrust fluctuations. Through Fourier analysis, the harmonic characteristics of the thrust fluctuation is achieved and the results are in good agreement with thrust fluctuation model. Finally, the positioning error of DSPMLSM is effectively reduced by adopting the thrust fluctuation suppression method proposed in this paper, which is of great significance to the further industrial application of DSPMLSM.

In the context of the energy revolution, the research and application of energy storage technology have been paid more and more attention. As one of the main energy storage devices, lithium-ion batteries are usually put into use in the form of battery packs in practice. Due to the inconsistency between battery cells, the battery pack cannot play the best performance. Moreover, there is a lack of sorting means in the recycling of battery cells. To solve this problem, a battery uniformity sorting method based on electrochemical impedance spectroscopy is presented. Seventy commercial batteries of the same type were sorted and grouped by this strategy. The capacity decay of the battery pack before and after sorting and the dispersion degree of the battery voltage curve were tested. It is proved that this method can achieve fast and effective sorting of batteries and has practical application prospects.

The four-wheel intelligent distributed drive vehicle (4WIDEVs) has been attracted a great attention in dynamics control due to its inherent actuation flexibility recently. However, frequent variation of vehicle speed in dynamic control is usually ignored or simplified. To enhance the driving safety in the extreme path tracking and obstacle avoidance maneuvers, this paper proposed an integrated dynamic control method. We first design the time-varying predictive model comprehensively considering velocity variation and yaw stability. Compared with the traditional speed constant model, by analyzing the mechanism of multi-degree of freedom nonlinear adaptive time-varying characteristics of four-wheel independent vehicles model, the influence of frequently changing speed on model accuracy is reduced. For ensuring the tracking accuracy, a linear adaptive time-varying predictive based control method (LATV-MPC) is developed to compute optimal front wheel steering angle and longitudinal tire force, where the predictive model is updated in each time horizon with the changing speed, eliminating the errors accumulation between the prediction model and real state. Simulation results based on MATLAB and CarsSim platform demonstrates that the proposed integrated dynamics control strategy allows the yaw stability to perform better and the tracking error to decrease in both double lane change and random obstacle avoidance scenario.

In view of the characteristics of distributed energy storage system with “large number and scattered distribution” of terminal devices, this paper proposes a star and chain two-layer networking mode. For devices with a long communication distance, they can connect to edge iot agent through sink nodes, and for devices with a short communication distance, they can directly communicate with edge iot agent. This networking mode can effectively enhance the reliability of system communication. According to the semantic rules of the Internet of Things, the iot terminal model of battery energy storage system is constructed from three aspects of static attributes, dynamic attributes and services, and it is combined with the traditional IEC61850 model. The IEC61850 model is responsible for the communication between edges, and the iot model is responsible for the communication between cloud edges. It not only solves the problem of information model heterogeneity between systems, but also breaks the situation of chimney system. Based on this, puts forward the MQTT protocol in the cloud - edge - end information interaction mechanism, from the themes of the communication architecture, interaction topic and protocol mapping specification design, makes the energy storage system has more standardized communication mechanism, effectively solve the current energy storage system terminal equipment communication protocol and networking mode difference big problem. Finally, taking an energy storage power plant system as an example, the MQTT client software is used to interact with the cloud for information, and the reliability and timeliness of this communication mechanism is verified through information interaction tests and time delay analysis.

In this paper, a layered control algorithm for energy-efficient predictive cruise control (PCC) algorithm of commercial vehicles that can optimize engine torque and gear in real time is proposed. The predictive cruise upper layer controller constructs a quadratic programming (QP) problem incorporating forward road information to optimize the vehicle driving force in the Model Predictive Control (MPC) framework. The predictive cruise lower layer controller optimizes the gear by finding the offline optimized gear MAP based on the overall vehicle driving force and current vehicle speed obtained from the upper layer controller. The predictive cruise algorithm is validated under highway conditions with slope changes. Compared with the traditional cruise control (CC) algorithm, it can save more than 4% fuel, and can better balance fuel economy, time economy, and driving comfort compared with the PCC algorithm based on indirect method solving. As for its solution speed, it is only twice as fast as the indirect method, which can meet the demand of real-time control.

Polyimide (PI) has excellent electrical insulation properties and is widely used in pulsed power devices and new energy vehicles. It is of great significance to study the breakdown characteristics and mechanism of insulating materials to ensure the safe and reliable operation of electrical equipment. The key problem is to establish the correlation between the spatial dispersion of polyimide charge transport parameters and the breakdown probability distribution characteristics. In this paper, polarization, trap distribution, carrier mobility and breakdown Weibull distribution of polyimide were studied. Then, considering the physical processes such as charge transport, molecular chain displacement and charge energy accumulation, the breakdown probability model of space charge and energy accumulation modulation was established. The breakdown probability distribution characteristics of polyimide were calculated by simulation, which follows Weibull distribution, and the relationship between characteristic breakdown strength and sample thickness follows inverse power function. The breakdown Weibull distribution curves with different shape parameters were obtained by adjusting the dispersion of charge transport parameters. By comparing the simulation results with the experimental results, the variances of four characteristic parameters of charge transport, such as carrier mobility, attempted escape frequency, trap energy level and trap density, are obtained by inversion. The dispersion of trap levels is a key factor affecting the shape parameters of breakdown Weibull distribution. The probability distribution of the breakdown strength is caused by the difference of charge trapping effect of traps in different spatial locations. With the increase of sample thickness, the dispersion of trap energy level increases, and the Weibull distribution shape parameter decreases gradually. The correlation between the spatial dispersion of trap energy levels and the breakdown probability distribution is revealed, which provides theoretical basis and model support for the design and risk assessment of high voltage power equipment.

Precise state of health (SOH) estimation is pivotal for reliable operations of lithium-ion batteries in electric vehicles. However, the collected charging data is usually incomplete, which makes it difficult to generate health features and brings great challenges to SOH estimation. To conquer this defect, Gaussian process regression (GPR) is developed using random partial charge segment to estimate battery SOH. Firstly, a voltage fitting process is proposed to reconstruct the constant charging voltage trajectories from random partial charging data. Then, the charging time is inferred to characterize battery deterioration. Correlation analysis is conducted and high correlation between SOH and health feature is verified. Following this endeavor, GPR is presented to effectively predict SOH with the input of charging duration. The proposed method can extend the random partial charging segment to complete charging data, thereby relieving the pain there is little chance that the drivers charge the battery from a predefined voltage data. Train and validation are executed on four battery cells, highlighting that the developed approach can maintain the SOH within 2% error.

Accurate capacity estimation is the cornerstone of attaining the state of health and remaining useful life of lithium-ion batteries. However, most of existing methods for battery capacity estimation are developed based on the fully charging/discharging condition, which is limited for onboard applications. This paper proposes a capacity estimation method based on an optimal voltage section. Firstly, the feasibility of capacity estimation based on sectional voltage data is demonstrated by correlation analysis between the voltage section-based health factors and the complete capacity. Secondly, the quantum particle swarm optimization algorithm is employed to determine the optimal voltage section. Thirdly, a mapping model between health factors and battery capacity is constructed using a long short-term memory neural network. Finally, validation results on public data sets show that the proposed method can realize accurate capacity estimation with an average root mean square error of 1.53%.

The widespread applications of supercapacitors are greatly restricted by their inferior energy density. Developing advanced electrode materials with high capacitance and large voltage windows has been regarded as a promising way to conquer the above challenge. Considering this, VO2@NiO core-shell heterostructures were constructed by a facile and controllable hydrothermal reaction combined with the atomic layer deposition technique. The NiO coating layer can buffer the hydrolyzed reaction of VO2, improving the chemical/electrochemical stability of VO2. Moreover, heterogeneous interfaces between NiO and VO2 can effectively regulate the charge distribution around the phase boundaries, which can render additional active sites and enhance the intrinsic electronic conductivity of two building blocks. as a result, the as-prepared VO2@NiO heterostructure exhibits extremely high specific capacitance of 1265 F/g at 1 A/g and remains 80.6% of initial capacity after 5000 cycles at 10 A/g. The assembled asymmetric supercapacitors employing the VO2@NiO heterostructure materials as the negative electrode and the commercial nickel-cobalt-aluminum oxides as the positive electrode exhibit an attractive energy density (39.81 Wh/kg) and superior cycling lifespans. This work may guide the development of core-shell heterostructure as advanced electrodes for electrochemical energy storage.

The solid-state diffusion coefficient is an important parameter to characterize the kinetics performance of lithium-ion batteries. It is the basis for establishing accurate electrochemical models. In this paper, the solid-phase diffusion coefficients of positive and negative electrodes of ternary lithium-ion batteries were measured based on electrochemical impedance spectroscopy. The expressions of battery diffusion process were derived by Fick's law, and then the impedance model of lithium-ion batteries was built. A genetic algorithm was used to fit the electrochemical impedance spectrum of lithium ion batteries. The solid-phase diffusion coefficients were obtained according to the time constant of diffusion process. The results show that the impedance model fits well with the experimental spectroscopy. At room temperature, the solid-phase diffusion coefficient of the positive electrode of the ternary lithium-ion battery increases first and then decreases with further delithiation of the electrode. The solid-phase diffusion coefficient is approximately 10–15 ~ 10–13 m2/s. In contrast, the degree of lithium intercalation has little effect on the solid phase diffusion coefficient of negative electrode graphite, which maintains about 10–13 m2/s. Traditionally, the Warburg impedance is used to measure the solid-phase diffusion coefficient only in the case of semi-infinite diffusion. In comparison, the proposed method in this paper takes varied diffusion conditions in consideration. It is more suitable for different types of materials, and has better application prospects.

Efficient and accurate prediction of battery remaining capacity can guarantee the safety and reliability of electric vehicles (EVs). However, battery capacity is difficult to measure directly due to complex application scenarios and sophisticated internal physicochemical reactions. This study develops a hybrid deep learning approach for accurate remaining capacity estimation based on differential temperature (DT) curve. First, the cycle life data are acquired and analyzed. Then, DT curves are deduced based on the charging data and smoothed via Kalman filter (KF). Next, health features (HFs) that characterize the battery degradation are excavated from the DT curves. Finally, a hybrid deep learning model fusion convolutional neural network (CNN) and gated recurrent unit (GRU) recurrent neural network (RNN) is established to predict battery remaining capacity. Each deep neural network (NN) in the model is engaged to execute a particular part in the forecasting task to maximize its corresponding merits. The superiority of the proposed method in terms of accuracy is justified via comparison with other modern methods including long short-term memory (LSTM) RNN, GRU RNN and a hybrid model integrating CNN and LSTM RNN. Experimental results demonstrate that the effectiveness and applicability of the proposed method in enabling battery remaining capacity estimation.

In the development of fuel cell hybrid vehicles, energy management strategy (EMS) plays an important role. The traditional EMS studies focus on the optimization of energy distribution in order to minimize the hydrogen consumption and degradation cost of fuel cell. However, the phenomenon of fuel cell recoverable performance loss which can affect the efficiency of fuel cell is ignored. Therefore, it is necessary to consider recoverable performance loss in the EMS of fuel cell hybrid power system, which can better simulate practical fuel cell hybrid vehicle scenarios. Although recoverable performance loss can be reversed by performing recovery procedures, fuel cell operation is interrupted during the procedures, thus it is worth studying appropriate time to perform recovery procedures without affecting total power output. To solve these problems, this paper proposed an EMS framework by considering fuel cell recoverable performance. In the study, fuel cell recoverable performance loss is converted to equivalent hydrogen consumption, then an advanced deep reinforcement learning method, Deep Deterministic Policy Gradient (DDPG), is selected to obtain EMS and determine the optimal time for conducting recovery procedure. Results show that with the proposed method, the total cost can be reduced by about 10% compared to EMS without considering fuel cell recoverable performance.

Transportation is one of the critical factors that leads to energy crisis, and over the years there has been a lot of research around the improvement of vehicle energy efficiency. To effectively reduce the overall energy consumption of continuous traffic flow in a road network, a collaborative eco-routing optimization strategy is proposed. The strategy takes advantage of the differences in energy consumption characteristics among vehicles to coordinate the path allocation among vehicles. It improves the energy consumption of low-energy-consumption vehicles and reduces the energy consumption of high-energy-consumption vehicles more significantly, thus achieving the effect of improving the overall economy. The simulation results show that the proposed method can effectively improve the economy of all vehicles in the road network by up to 4.85% without additional time cost.

The accuracy and robustness of the state of charge (SOC) estimation play an important role in the overall deployment and control of the battery management system. To maintain high accuracy and stability of SOC estimation in the whole range, the author uses OWA (ordered weighted average) operator to fuse three Kalman filter algorithms. OWA operators can assign weights to different algorithms based on values that approximately characterize the estimation accuracy. In this paper, the weights are updated based on the voltage residual and the real covariance respectively. The results show that whether it is the OWA fusion algorithm based on voltage residues or the OWA fusion algorithm based on the real Covariance, the estimated accuracy and robustness are better than the single algorithm.

A large number of small-capacity distributed energy storage (ES) systems have been introduced to take an important part in grid frequency regulation. However, the accompanying high-order optimization problem causes an inevitable issue for both centralized and distributed control methods. Accordingly, based on the reduced-order predictive model, a two-layer control strategy of frequency support for power systems is proposed in this paper, in which the frequency regulation and economic dispatch are combined. The lower layer is implemented in a distributed framework which only requires communication between adjacent ESs, and the upper layer only utilizes the reduced-order information sent from the lower layer which lessens the computation burdens. The effectiveness of the entire control method is validated under different scenarios by simulations. As indicated in the simulation results, the proposed method achieves almost the same control effect in ES clusters as that using centralized control with a shorter computing time. And the proposed method incorporates the widespread multiple small-capacity ESs effectively for power system frequency control.

The operating characteristics of the continuous climbing time, the long downhill retarding distance, and the wide load variation range make the fuel consumption of mining dump trucks high, and the energy consumption cost can reach 1/3 of the total operating cost. The electric retarding technology with the resistance cabinet is the only way to dissipate the braking energy of large mining dump trucks. However, it cannot realize the recovery and reuse of braking energy. High instantaneous braking power, short braking time, less energy recovery, and short life of the energy storage components make the lithium battery an auxiliary energy storage hybrid technology unable to effectively adjust the engine load rate to achieve the expected energy saving and cost reduction effect. Therefore, a typical operating condition model is established based on the operational characteristics of open-pit mines. Combined with the three topology structures of motor light overload, cooling fan electrification, and reverse drag engine, four technical solutions are proposed to improve the energy utilization rate of mining dump trucks. “Energy-saving coefficient of feedback braking energy” is the evaluation index, taking a mine dump truck with a load of 150 tons as the scheme’s application object and comparing each scheme’s energy-saving effects. The results show that on the premise of ensuring the vehicle’s dynamic performance, maximizing the use of braking energy can reduce engine energy consumption by 2–11%.

The integrated energy system (IES) has seen widespread application in the energy production as a result of the advancement of energy intelligent technology. While significantly increasing energy efficiency, small hole leakage failures of gas pipelines bring operational risks to the energy supply. Gas pipeline leakage may cause explosions, threatening people and buildings. Therefore, based on the one-dimension gas flow equation and dynamic gas flow model, a system network model that considers the change of pipeline topology caused by leakage failure is proposed. The process of gas leakage in the operation of gas pipelines is illustrated in the system network model. Then, an optimal scheduling model for IES is established, which includes the dynamic characteristics of gas system and leakage failure. The integrated gas and electricity testing systems are used to validate the suggested model.

New energy vehicle fire accidents have raised concerns about their safety in recent years. Two indicators of maximum temperature and temperature extreme difference, which are closely related to heat, were chosen for the thermal safety of new energy vehicles. The rate of heat transfer in actual, everyday use, is, however, significantly lower than the comparable rate of voltage. As a result, indicators that express the consistency of the battery pack include voltage polarity and voltage information entropy associated with voltage. The thresholds for the four indicators were then set using a study of the literature and $$3\sigma$$ 3 σ different techniques, and voltage clustering was performed to check for potentially dangerous single cells.

Battery is the key component and main trouble source in an electric vehicle (EV). With rapid growth of market share, thermal runaway caused by malfunction of batteries have been frequently reported, making fault diagnosis essential to ensure safety as well as to promote performance. Unfortunately, most of existing fault diagnosis methods focus solely on identification of abnormal single cell while ignoring characteristics of battery macro system, failing to catch error of certain types. In this paper, a novel fault diagnosis method based on Kolmogorov complexity theory is proposed and verified by the real vehicle fault data from China’s national big data monitoring platform for electric vehicles. On top of Kolmogorov complexity, degree of confusion inside a battery pack can be quantitatively described, where analysis results clearly show the correlation between increase of confusion and thermal runaway accident. As a brief conclusion, the proposed method can be an important complement to traditional methods.

Hybrid energy storage systems (HESSs) are playing an increasingly important role in smart mobility platforms including electric vehicles. The design of the energy management strategy is the core of making the system rationalize the power distribution and stable operation. The power state and temperature state directly affect the determination of safe operating boundaries in the energy management strategy. In this paper, based on the equivalent circuit model of HESS, a thermoelectric coupling model of battery pack considering air-cooled system is established. In addition, a power capability prediction method considering the constraints of temperature, current, voltage and SOC is proposed and the power capability prediction method is embedded in the energy management strategy based on model predictive control. The experimental results show that the proposed algorithm can further improve the system performance and reduce the system energy consumption. The energy management strategy can provide the optimal power distribution at different air-cooled wind speeds and guarantee the maximum temperature of both the battery pack and the supercapacitor pack are in the normal operating range.

The community microgrid is a potential solution to integrate distributed energy resources into the main grid. The Stackelberg game is widely used to solve the energy dispatch problem for the community microgrid, where the utility functions of prosumers have to be publicly available. However, such a scheduling strategy is challenged due to the increased prosumer privacy awareness and the difficulty of accurately describing the utility. In this paper, we propose a signaling game-based energy scheduling strategy for a community microgrid with shared energy storage. The utility functions of prosumers are private and unavailable to others. An improved Bayesian optimization algorithm (BOA) with a sliding window is proposed for game equilibrium acquisition. In addition, the existence and uniqueness of the game equilibrium solution are proven, and the effectiveness of the BOA algorithm is demonstrated. The superiority of the proposed method is verified by simulation experiments using the Stackelberg game as the benchmark. The utility of the shared energy storage provider is increased by 17.4%, while the utilities of the prosumers are guaranteed to remain almost constant. In addition, the impact of the community microgrid on the main grid is reduced by approximately 20%.

The quality of the Lithium-ion battery charging performance straight affects clients’ awareness and acceptance of the electric vehicles. Researching on the optimization of the charging methods is critical for the evolution of more intelligent battery management systems and smart electric vehicles in the future. This paper first introduces the types of simple charging methods and the existing shortcomings, and then describes the characteristics of various optimized charging methods and compares them, then proposes a fast charging strategy based on the DDPG (deep deterministic policy gradient) algorithm. By balancing the charging speed, battery life and safety, the minimum penalty target function is established by Zhong et al. 1699–1704, 2004. Combining the model-based state observer with the general algorithm framework of the optimizer based on deep reinforcement learning, a context-aware deep deterministic policy gradient algorithm with priority experience playback is proposed. Compared with the traditional charging methods, the optimized charging method can shorten the charging time, improve the charging performance, prolong the battery cycle life, and perform a clever compromise between the charging speed and the compliance with physical constraints.

This chapter presents a differential drive based cooperate steering control strategy considering energy efficiency for multi-axle distributed drive vehicle. Apart from used as a fault-tolerant control method of Steer-by-Wire (SBW) system to keep stability and performance of the vehicle, the differential drive steer (DDS) can also be combined with SBW by reasonable torque allocation with considering the efficiency distribution of drive motors during normal driving. Firstly, the dynamic models of multi-axle vehicle including the front SBW system are established. Then, the upper layer path follow controller is designed based on model predictive control (MPC) with measurable disturbance to obtain the generalized steer torque. According the drive motor efficiency map, the lower layer steer axles optimal torque allocation strategy is proposed to balance the contribution of SBW motor torque and differential torque. Finally, the strategy is validated through close-loop simulation in MATLAB/Simulink. The result shows that the proposed strategy can adaptively adjust differential torque of steer axles and SBW motor torque to track the target path with better energy efficiency.

With the application of high-capacity lithium iron phosphate (LiFePO4) batteries in electric vehicles and energy storage stations, it is essential to estimate battery real-time state for management in real operations. LiFePO4 batteries demonstrate differences in open circuit voltage (OCV) under different charge and discharge paths, indicating the hysteresis phenomenon of OCV, which is more evident under energy storage frequency regulation conditions. Previous battery models ignored the hysteresis characteristics in the energy storage frequency regulation conditions, causing low accuracy in the state of charge (SOC) estimation. To accurately estimate the SOC of LiFePO4 batteries, a hysteresis voltage reconstruction model is developed to analyze the hysteresis characteristics of LiFePO4 batteries under automotive dynamic conditions and energy storage frequency regulation conditions. The accuracy of the hysteresis model is compared with the basic first-order RC equivalent circuit model. Furthermore, the SOC estimation based on the extended Kalman filter (EKF) method is achieved. Results indicate that the hysteresis model exhibits better accuracy for the hysteresis features, with an error of less than 1.5%, which is more appropriate for SOC estimation under energy storage conditions.

The fractional-order model is effective in simulating the battery’s dynamic behaviour. However, the real-time implementation of a fractional-order component can be complicated as the related Grünwald–Letnikov (G–L) definition involves sophisticated cumulative operations. Here, we propose to use an integer-order model with frequency-dependent parameters to approximate the fractional order model. Its implementation is described as follows: In the frequency domain, multiple third-order RC models are switched with frequency to depict the battery’s impedance spectroscopy. In the time domain, our model is transformed into a weighted summation of multiple third-order RC models, whose weighting factors are determined by the current signal’s energy in the frequency ranges of interest. The comparison between the integer order RC model, the proposed frequency-dependent RC model, and the fractional order model is carried out in both the time and frequency domain to verify the effectiveness of the proposed method. The computational effort of the proposed model can be significantly reduced by 85% compared with the fractional order model, and the modelling error is reduced by 51% compared with the conventional integer order model. Our model provides an accurate yet computationally efficient way to describe the battery’s dynamic.

The high proportion of renewable energy connected to the power grid leads to insufficient regulation capacity. Physical energy storage system can provide fast regulating capacity, but, it is seldom used on a large scale in power system due to its cost. The virtual energy storage system which aggregates a variety of flexible load resources can also achieve the same effect as physical energy storage. The scheduling of virtual energy storage depends on the accurate prediction of its power baseline. This paper analyzes the multi-dimensional factors that affect the baseline of virtual energy storage elements, including temperature, date attributes and electricity price. Considering the above factors, an adaptive baseline prediction method based on BP neural network, SVR and LSTM neural network algorithm is designed. Numerical analysis shows that the error of the proposed method is only about 20% of that of the method actually used in the demand response of a city in China.

In order to safely and efficiently use their power as well as to extend the life of Li-ion batteries, battery fault diagnosis is essential for ensuring the safe and reliable operation of electric vehicles (EVs). However, there are still many difficulties in solving this problem. First of all, we need to do analysis and visual analysis based on the original data of the battery. Secondly, because the data of the power battery is time-series, and the occurrence of battery failure has the characteristics of evolution with time, to solve this problem, this paper proposes a time-series prediction method of SOC based on the long short-term memory recurrent neural network (LSTM). According to the time order of the collected data and the long-term dependence of LSTM learning, the analyzed battery data is divided into the train set and test set, and the battery SOC value is predicted by establishing a long short-term memory recurrent neural network model for training. Finally, calculate the accuracy, error and fitting degree. The experimental results show that the proposed method can accurately predict the SOC of the power battery.

Thermal runaway accidents of lithium-ion batteries have occurred frequently in recent years, which have seriously affected its further promotion and application. In the early stage of battery faults, its external fault performance is usually relatively minor. And battery packs are often accompanied by inconsistencies, which further increases the difficulty of effective identification of faulty cells. This article proposes a battery fault diagnosis method combining variational mode decomposition (VMD) and edit distance. Combined with a thermal runaway case in actual profiles, VMD is adopted to decompose discharging voltage signals into static components. And the edit distances between all cells are extracted from the charging curves and discharging static components in certain voltage intervals. Finally, the unsupervised Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm is implemented to identify the faulty battery in the battery pack. And the voltage abnormality is detected 10 days before thermal runaway, to realize the warning of battery thermal runaway fault.

Aiming at the problem that the change of capacity during the aging process of lithium-ion batteries affect the accurate estimation of state-of-charge (SOC) and state-of-health (SOH), this paper proposes a joint estimation method combines Ordinary Least Squares (OLS) and Unscented Kalman Filter (UKF) algorithm. First, OLS algorithm is used to estimate SOH online to improve the prior accuracy of SOC. Then, the SOC is estimated by UKF algorithm. The experimental results indicate that the joint SOC-SOH algorithm can realize the accurate estimation of SOC and SOH during battery aging. The SOH estimation error is within 1.5%, and the SOC estimation error is within 2%.