Research On SOC Estimation And Overcharge And Overdischarge Fault Identification Of Electric Vehicle Lithium Battery

Due to the exhaust emissions of fuel vehicles and the impact of fossil fuel consumption on the earth’s environment and resources,electric vehicles began to enter people’s lives.The main advantages of electric vehicles are: reducing people’s dependence on fossil energy,reducing exhaust emissions,low operating cost,fast response and so on.Electric vehicles will have a prominent position in the future automotive market.As the core of the battery management system,the accurate estimation of battery power plays an important role in extending the endurance and service life of electric vehicles and preventing the overcharge and over discharge of single battery.However,in the battery pack,due to the inconsistency between each cell,individual cells will experience overcharge and discharge after several charge discharge cycles due to the inconsistency.Therefore,accurate identification of overcharge and over discharge fault battery is very important to ensure the safe operation of electric vehicles.Combined with the above two key contents,this paper will carry out the accurate estimation of the state of charge and the identification of overcharge and overdischarge fault battery of electric vehicle lithium battery.Firstly,the advantages and disadvantages of several lithium batteries are compared,and the internal reaction principle of lithium iron phosphate battery during charging and discharging is introduced.The main characteristics of lithium battery are introduced and analyzed.Secondly,the off-line parameter identification of the selected lithium battery is carried out.According to the state equation and observation equation of the selected second-order Thevenin equivalent model,the simulation model is established in Simulink,and the accuracy of the model is verified through a variety of working conditions.According to the algorithms of Kalman,extended Kalman and dual Kalman filter,combined with the battery model mentioned above,the simulation models of extended Kalman estimation SOC and dual Kalman estimation SOC are built.Through a variety of dynamic and non dynamic conditions,the accuracy and convergence of the two models are verified,and it is proved that the accuracy of dual Kalman estimation SOC is better than that of extended Kalman algorithm.Finally,the parameters of overcharge and overdischarge battery are extracted and brought into the established Kalman filter to build a multi model estimation simulation model.The matching degree between the current battery and the model is judged by the residual and weight of the model output representing different battery fault states,and the overcharge and overdischarge fault identification of lithium battery is completed.