At low temperatures， the lithium-ion batteries of electric vehicles have the problems of usable capacity reduction， charging difficulty and cycle life decay， which seriously restrict the application of lithium-ion batteries. Therefore， it is very important to ensure that the lithium-ion batteries operate within the appropriate temperature range. With the advantages of fast heating rate， good temperature uniformity and simple system structure， the battery pulse heating technology is an effective method to solve the problem of low temperature application of the lithium-ion batteries. In this paper， the research progress of pulse heating technology is summarized from the three aspects of pulse heating schemes， pulse control parameters and pulse heating strategies. Firstly， the advantages and disadvantages of the existing pulse heating schemes are introduced. Secondly， the temperature rise and capacity decay characteristics of lithium-ion batteries under different pulse control parameters are summarized. Finally， the influence of different pulse heating strategies on the low-temperature performance of lithium-ion batteries is compared， and the future development direction of pulse heating technology is pointed out.
The prediction of remaining useful life （RUL） of lithium-ion power battery is of great significance for understanding the safety and reliability of electric vehicles in the whole life cycle and improving the design of battery management system. Generally， the time series prediction method based on deep learning is a recursive process. The error of each prediction will accumulate with the increase of prediction times， which is difficult to ensure the prediction accuracy and efficiency. Based on the theory of deep learning sequence prediction and error analysis， an ARIMA-EDLSTM fusion model is established for lithium battery remaining useful life prediction. The encoder decoder （ED） framework is used to improve the long short-term memory neural network model （LSTM）， establish the EDLSTM model of sequence to sequence prediction， and fuse the ARIMA model to predict the error trend and modify the prediction results. Theoretical analysis and real vehicle data verification show that this method can still better fit the real vehicle SOH decline curve when the prediction proportion exceeds 35% of the total history data， and effectively improve the prediction accuracy of the remaining useful life of lithium battery.
As a key parameter of battery module and system thermal design， the specific heat capacity of lithium-ion batteries has a significant impact on the accuracy of the simulation model and control strategies of the established battery thermal management system. Based on the calorimetric method， a test platform with low cost， simple implementation and high measurement accuracy is established. The influence of heat loss， calorimetric method and hierarchical optimization of the calorimetric device on the test results is systematically studied. In the sample calibration test of brass， 304 stainless steels， cast iron and high-density polyethylene standard parts， the experimental results show that the specific heat capacity error of the calorimetric device is less than 3%， and it is not affected by the thermal conductivity of the test sample. The average specific heat capacity of 32650 LiFePO4 batteries is determined to be 1.022 J·g-1·K-1 by this experimental device， which provides a feasible method to accurately measure the thermal physical parameters of specific heat capacity of other types of batteries and even next-generation solid-state batteries.
At present， there is no effective method to identify and detect the micro short circuit of lithium-ion battery in the early evolution stage. Therefore， this paper proposes a micro short circuit diagnosis method based on the change law of battery charge capacity increment （IC） curve and charge capacity difference （DCC）. Firstly， the relationship between lithium battery short circuit fault and charge capacity increment is established. The IC curves are denoised by using wavelet transform， and the unique correspondence between the peak value of IC curve （ICPV） and the state of charge （SOC） of battery is obtained under different current rates and temperature. Then， DCC is proposed to be used to describe the SOC difference between the faulty battery with internal short circuit and the normal battery， and the quantitative method of lithium battery micro short circuit is obtained. Finally， simulation analysis and experiments show that the quantitative information of battery micro short circuit can be obtained under different cycling conditions， and the maximum errors of fault diagnosis are less than 8.12%.
Based on the complex operation data of real-vehicle batteries， the IC peak features are extracted as effective features of battery charging segments using incremental capacity analysis method in this paper， and the IC peak features are processed using t-SNE nonlinear dimensionality reduction method to eliminate the redundancy of multidimensional features and solve the problem that it is difficult to extract reliable features from real-vehicle data. In addition， a support vector regression model is constructed to estimate the battery health status. The results show that the use of incremental capacity curve peak features can effectively characterize the recession changes of the battery health state. The smoothing and noise reduction methods for real vehicle data can improve the quality of training data better. The SVR model based on t-SNE dimensionality reduction features improves the estimation accuracy of battery health state and ensures accurate estimation on a limited sample data set.
In this paper， a battery thermal management system based on air cooling coupled with the composite phase change materials cooling （referred to as APE-BTMS） is proposed， in which the composite phase change materials （PA/ EG） are obtained by mixing the paraffin wax （PA） and the expanded graphite （EG）. In this system， the middle part of the battery is cooled by PA/ EG， and the upper and lower ends of the battery are cooled by air with the air velocity of 1.23 m/s. The main purpose of APE-BTMS is to reduce the total weight of the battery thermal management system while cooling the working temperature of the battery to the optimal temperature range. The experimental results show that the APE-BTMS-45 model has the best cooling performance under the same conditions. A numerical model of APE-BTMS based on COMSOL is established to compare the cooling performance of APE-BTMS at different ambient temperatures and more finely axial thickness. The numerical simulation results further verify that APE-BTMS-45 has the best cooling performance in the comparative data， and can reduce the weight by 216.71 kg at maximum. The research results of this paper can provide reference and data support for the design and development of battery thermal management system （BTMS） based on phase change materials.
The semi-empirical model of the proton exchange membrane fuel cell has several unknown empirical parameters with widely varying scales and ranges of values， which can lead to biased results due to the loss of local high sensitivity information in the global sensitivity analysis. For this reason， double cost functions are constructed that are reciprocal to each other. On the basis of the original cost function， the correlation between the empirical parameters and the response error in the full domain is calculated. A first global sensitivity analysis is performed using the Sobol method for uniform sampling. The parameters with high sensitivity （i.e. affecting the speed of convergence） in the whole range of values are filtered； then the reciprocal cost function is used to amplify the error correlation in the local area， and for the remaining globally insensitive parameters， another global sensitivity analysis of the reciprocal cost function is carried out to filter the parameters with high sensitivity （i.e. affecting the accuracy of convergence） in the local range of values. Thus， the ability of identification of multi-scale and multi-local high-sensitivity parameters is improved. The analysis results show that the response error of the model after high-sensitivity parameter identification is consistent with the results of the full parameter identification， saving about 60% of the computational cost. The applicability and accuracy of the method are verified by fuel cell stack experiments.
Based on the OpenFOAM platform， a 3-D gas diffusion layer （GDL） reconstruction method is developed independently in the paper， and a numerical model is established by employing the dynamic mesh method to investigate the water transport in GDL of fuel cell under various vibration conditions， including vibration directions， frequencies and amplitudes. The results show that compared to the vibration in the horizontal direction， the impact of the vibration in the vertical direction on water transport is more significant. The water saturation presents regular sinusoidal vibration characteristics in high-frequency vibration， but the periodic characteristics of water saturation in low-frequency vibration are irregular or not significant， which is close to sinusoidal change on the whole. This study has certain guiding significance for the fuel cell stack layout and shock absorption design.
CTB （cell to body） battery baody integration technology has great advantages in improving endurance mileage， vehicle stiffness and crashworthiness， which has become a new development direction of the new energy vehicle industry. However， sealing is one of the biggest problems that restrict the development of CTB technology in order to integrate the upper cover of the battery pack with the floor of the vehicle body. So far， there is still no research in the field of CTB sealing in the industry. This paper carries out studies in terms of CTB sealing strategy， sealing structure design， sealing component selection， failure consequence analysis and user condition design verification， and puts forward solutions to the CTB sealing design problems in the industry for the first time， so as to accelerate the popularization and application of CTB technology， and promote the electrification transformation of the global new energy vehicle industry.
The torsional stiffness of the body-in-white is an important mechanical performance index of the load-bearing body， which has a direct impact on the handling stability of the vehicle， and is also an important index to evaluate the lightweight level of the body. Since the new generation of pure electric body has a large structural frame difference from the traditional body， in the early stage of body design， the force transmission path is redefined with the goal of improving torsional stiffness. In this paper， based on theoretical analysis and topology optimization， the optimal force transmission path of the torsional stiffness of the body is found. Through the integrated design of the battery pack and the body， the body forms a plurality of closed structures that are close to a circular ring. On the premise of no extra weight adding to the car body t， the torsional stiffness of the body-in-white can reach 40，000 N·m/（°）， and the lightweight coefficient of the body can reach 1.75， achieving the leading level of pure electric vehicles， and at the same time greatly improving the handling stability of the vehicle.
The platformization development of vehicles and the corresponding systems is the consensus and important technology trend of the new energy vehicle industry， which is beneficial for planning each project systematically， sharing more systems and saving the cost. It is necessary for platformization development of battery systems based on vehicle platforms to clarify the relevant variables of vehicles and batteries， which mainly include the space， the weight， the mileage， the power consumption， the power performance of the vehicle， and the size， the weight， the energy， the power property of the battery. In this paper， the development of the battery system of a certain hybrid-vehicle platform is chosen as the case， where the platform solution including two or three types of cells is proposed via summarizing the distribution and boundary of each variable. Moreover， the bandwidth design of the series of the cells and the Y dimension of the pack is also proposed to make it compatible with 16 vehicle types. This study provides the basic reference of strategies for the development of battery solutions for various vehicle platforms of the industry in the future.
In order to accurately predict the battery safety risk， a multi-index battery safety risk prediction method based on vehicle-weather-driver is proposed in this paper. Firstly， multi-dimensional information inside and outside the vehicle is extracted， i.e. multi-index characteristics such as weather condition， driving conditions and driving style are extracted by data mining to simulate the actual battery application scenario. Then， features are filtered by random forest and SHAP combination model， which improves the generalization and robustness of the model. Finally， the battery safety risk prediction problem is decoupled into machine learning prediction and time series prediction problems， and XGBoost and random forest models are selected to predict respectively. On this basis， a new Stacking integrated model is established to predict the battery safety risk. According to the predictive effect of the final model and the results of data experiment， the scheme can make a more accurate prediction of the battery safety risk of EV and provide decision-making information for safe and intelligent battery management system.
In order to improve the prediction accuracy of remaining useful life of lithium-ion power battery in practical application， a remaining useful life prediction method of lithium-ion power battery combining the empirical aging model and the battery mechanism model is proposed in this paper. The method uses the SOH prediction value based on the empirical aging model as the prior estimate of the Kalman algorithm， and uses the SOH predicted by estimating the future capacity decline of the battery based on the mechanism model as the posterior correction of the Kalman algorithm， so as to achieve accurate prediction of the remaining useful life of the lithium-ion battery. The validation results of power battery remaining useful life prediction algorithm based on the cell test data show that the remaining useful life prediction error of lithium ion power battery is ≤ 5.83% and the maximum error of remaining useful life prediction of lithium-ion power battery based on real vehicle data is 8.12%， which has achieved good prediction results and enriched the life prediction methods of lithium ion power battery.
Battery safety has become an important issue restricting the development of electric vehicles. Accurate and timely battery safety warning can ensure the safety of occupants' life and property and improve the safety level of electric vehicles. A comprehensive review on the battery safety warning strategies in electric vehicles is conducted in this paper. Firstly， the definition of battery safety state is reviewed， and the framework of this review is proposed. Then， the battery safety characteristics and safety influencing factors analysis， battery-modeling methods， battery safety risk assessment/prediction methods are reviewed in detail. The advantages and disadvantages of different methods are summarized. Finally， the achievements and deficiencies of existing researches are summarized， and the development trend of battery safety warning technology in electric vehicle is proposed， including the new sensor technology， multi-factor integrated battery safety warning method and "terminal-side-cloud " battery safety warning system. This review provides a reference for further research on the safety warning strategy of electric vehicle battery.
At present， there is no effective method for unsupervised fault warning for vehicle cloud data with unspecified fault types. Therefore， this paper proposes a multi-level fault warning method for lithium-ion batteries driven by cloud data. Firstly， the features suitable for the characteristics of cloud data are selected through mechanism analysis， and six types of differential entropy feature sets are constructed for multiple mixed clustering to achieve the score evaluation of battery health. Then， temperature information is introduced in to distinguish heat-related faults and the warning level division criteria are constructed to determine the battery fault status. Finally， five field failure cases are used for validation. The results show that the method can accurately identify faults and distinguish fault types， and is ahead of its time and highly adaptable.
Due to the existence of hysteresis characteristics， it is difficult for battery management systems to accurately obtain the state relationship between open circuit voltage （OCV） and state of charge （SOC）. In order to effectively operate and manage the power battery， this paper investigates a lithium-ion battery model that considers the hysteresis characteristics and selects FFRLS for online identification of parameters. A SOC estimation method combining gated recurrent unit （GRU） neural network and adaptive extended Kalman filter （AEKF） is proposed， using the estimated results of the AEKF and GRU neural network as the model and measured values respectively， and the final SOC estimation results are obtained by Kalman filter （KF） ， which is used as the input to the AEKF at the next moment. The results show that the root mean square error （RMSE） of the prediction of voltages by models considering hysteresis characteristics and the SOC estimation by the joint estimation method is within 0.5 mV and 0.64% respectively for the ambient temperature environment. The RMSE for terminal voltage prediction and SOC estimation is within 0.9 mV and 0.72% for low and variable temperature environment respectively. The model considering the hysteresis characteristics and joint estimation method have good accuracy and robustness.
The fuel cell system includes stack， air subsystem， hydrogen subsystem， cooling subsystem， involving many components. Therefore， in the early stage of research and development， establishment of a fuel cell system model through system simulation has a guiding role for system development. Firstly， based on the test results and characteristic parameters of components， virtual calibration of components is carried out to establish an accurate component model. Then， according to the system flow chart， a complete fuel cell system simulation model is built. Finally， key output performance parameters of the system are evaluated and predicted through simulation calculation. Comparing the simulation results with the test data， the results show that the maximum average absolute percentage error between the model simulation results and the test data is 4.33%， with a high degree of consistency. It is verified that the simulation model has a high accuracy and can be used to study the performance of the fuel cell system， which has great guiding significance for future research and development of the fuel cell system.
In order to improve the uniformity of temperature and reactants between different cells in the stack， thus improving the stack performance and extending the stack life， an improved scheme of setting different gradient porosity in gas diffusion layers of different cell units in the stack is proposed in this paper. A three-dimensional， non-isothermal， single-phase short stack model with five layers of cell units is established for analysis. It is found that the scheme with porosity of 0.4-0.5-0.6-0.5-0.4 can minimize the difference of temperature， oxygen， water molar concentration， membrane current density between the edge layer and the intermediate layer cell unit； thereby improve the internal uniformity of the stack， which is also the same trend under the operating conditions of gas shortage.
Accurate and efficient abnormal detection of electric vehicle power battery systems is of great significance to ensure safe and reliable operation of vehicles. Based on this， a new power battery voltage abnormality diagnosis method based on voltage variation rate is proposed for detecting abnormal voltage fluctuation faults of individual cells in a battery pack. Further， an evaluation coefficient based on an improved Z-score method is introduced to quantitatively characterize the degree of abnormal voltage fluctuation. On this basis， the validity and reliability of the proposed method is verified based on real-world vehicle data. In addition， a comparative analysis with the commonly used entropy method shows that the method proposed in this paper has reliable fault diagnosis results and high calculation efficiency， with higher value of engineering application. Finally， based on the model， the distribution of the risk of voltage abnormalities in the battery system of this type of vehicle is obtained by statistically analyzing the voltage data of a large number of electric vehicles of the same type. By analyzing the abnormalities hidden beneath the surface， it can provide a reference for vehicle manufacturers for design of the power battery system or the entire vehicle structure.
For the research on safety risk management and control of new energy vehicle power batteries， this paper discusses in detail the failure mechanism and types of power battery systems， clarifies the coupling relationship between battery consistency and safety based on big data statistical analysis， and summarizes the data-driven safety state prediction， fault diagnosis and warning method. Finally， a "vehicle-cloud"-integration-based safety control strategy is proposed for real-vehicle battery systems. This paper aims to provide theoretical guidance for realizing real-time monitoring of battery safety status and risk warning for real vehicles.
