SOC Estimation Of Electric Vehicles Based On Span-lateral Inhibition Broad Learning

With the fast development of the electric vehicle(EV)business,research on the core key technologies of car battery that seriously influence and limit the energy performance of electric vehicles has become a front-burner topic.Among them,the state of charge(SOC)is an important parameter of the battery management system(BMS)of EV.It can be used to characterize the remaining power of the car battery and its working condition,in order to reasonably arrange the charging time and other related matters to ensure the reliable and safe work of EV.In the complex working conditions of EV,the current and voltage will change dramatically of the battery,and the temperature will also change,so estimating the battery SOC is a big challenge.Real-time monitoring of the battery system requires accurate battery SOC to effectively prevent battery damage caused by overcharging or overload.In this paper,the non-linear relationship between some factors that affect the estimation of the battery SOC and it is analyzed.The forecasting methods commonly used for the estimation of battery SOC are summarized.And we analyzed the characteristics of various SOC forecasting methods.To improve the prediction accuracy of battery SOC,this paper proposes a new method for battery SOC estimation using the Span-Lateral Inhibition Board Learning System(S-LIBLS).By analyzing the key factors affecting SOC,the input signal of the model is determined.The data are then generated in the laboratory by applying the driving cycle load to the lithium-ion battery at different ambient temperatures to ensure that the battery is exposed to a variable dynamic environment.S-LIBLS can timely encode dependencies into network weights and achieve accurate estimation of SOC in this process.In order to verify the performance of S-LIBLS,one sample training network from-20°C to 25°C was selected.The results showed that S-LIBLS could obtain accurate SOC estimation under various environmental temperature conditions.