State Of Health Estimation And Capacity Decline Prediction Of Lithium Ion Batteries For Electric Vehicles

In recent years,due to the energy crisis and the deteriorating environment,electric vehicles have become the research hotspot of major global automobile manufacturers.Lithium ion batteries are the power source of electric vehicles.In order to ensure the safe and efficient operation of electric vehicles,it is necessary to monitor the battery status in real time.Accurate estimation of the current state of health of batteries and prediction of the decline trend of capacity are conducive to the diagnosis of batteries and to the designation of reasonable maintenance and replacement strategies for lithium-ion batteries.In this paper,the following work has been done on the state of health estimation and capacity decline prediction of power lithium-ion batteries.Through the analysis of the internal structure,charging and discharging mechanism and the influencing factors of the state of health of the battery,the failure process and the change of the external characteristic parameters of the battery were studied.The capacity of lithium-ion batteries is determined as the evaluation index of the state of health of lithium-ion batteries.Based on the state parameters of the charging and discharging process of lithium-ion batteries,the combination of voltage,current,temperature and SOC as health factors is proposed to estimate SOH.On the basis of determining the health factors,using the excellent non-linear mapping ability of support vector machine,the relationship between health factors and state of health is established by support vector regression machine.Radial basis function is selected as the kernel function of support vector regression machine.According to particle swarm optimization and grid search algorithm,the optimal parameters of support vector regression model are obtained,and the model is trained.The results show that the selected health factors can accurately estimate the state of health of power lithium-ion batteries under dynamic conditions in MATLAB environment.In the actual use of power batteries,the decline trend of battery capacity is affected by many factors,so it is difficult to establish a reasonable decline model.In this paper,BP neural network is used to predict the future decline trend of batteries.To solve the problem that BP neural network is sensitive to the initial weight threshold,ant colony algorithm is used tooptimize the initial weight threshold of BP neural network.Finally,the battery cycle life test data from NASA's open data set is validated.The validation results under MATLAB environment show that the method based on ant colony optimization BP neural network can accurately predict the capacity decline trend of lithium ion batteries.