Research On Trchniques Of On-line Estimation Of State Parameters Of Lithium Ion Battery

Battery Management System(BMS)is one of the most important systems of electric vehicles,which all the batteries are managed by the battery management system.A good Battery Management System helps the electric cars to work safely.The state parameters of the battery which are the basis of the battery management system and other vehicle management systems provide the most basic parameters for other systems.This paper`s main content is about the SOC which describes the battery’s state of charge and SOH which depicts the life of battery.Accurate lithium battery electrochemical model is conducive to accurate the estimation of state parameters.In this paper,we focus on the lithium-ion battery for the electric cars.For the battery management system;we used the electrochemical mechanism of the battery,and used the electrochemical model to develop the state estimation algorithm.Because of the complex chemical reaction inside the cell and the complexity of the model,the electrochemical mechanism of the battery is simplified,and the simplified model is used to estimate the parameters on line.Firstly,we considered the traditional pseudo-two-dimensional model as the main model to estimate the parameters and then compared it with the Single Particle Model.To use it into the BMS,we analyzed the advantages and disadvanteges of these two models and combined it.Then,we proposed a new model which is improved for the single particle model.Finally,we chose the pseudo-two-dimensional model as standard and analysed if the new model could fit the requirement.Starting from the lithium battery model,the on-line estimation of the state parameters of the battery was studied which is based on the electrochemical model of lithium battery.Because the state of charge of the battery is an intuitive expression for the existing electric,researching the state of the charge is conducive for the battery management system to control the situation of batteries in the electric car.We chose the EKF and GA to analyse the charge of the battery which is describled by the modified algorithm of the single particle model,and the conclusion showed that the EFK and GA could estimate the SOC well.Becuase of the complexity of the Electrochemical model,the GA would be better in conclusion which`s deviation is less than 1%.Aimed at the common battery health estimation algorithms,we selected capacity as the basis of the health state of the battery management system.Meanwhile we analyzed the factors which might influence the capacity of the battery in the actual battery mechanism model.Considering the health of battery,we chose the Coupled-tank System to analyse the health of the battery and compared the conclusion with the conclusion of the circulation life test of battery in order to verify wether the algorithm is proposed.The conclusions show that the lithium insertion rate and the volume fractions of electrode material are the most important factors which influence the health of the battery.We combined the battery management system with the actual electric vehicle,which the estimation algorithm is applied to the actual project,and we compared the conclusion of experiment and simulation of the SOC and SOH,to verify the feasibility of algorithm,and finded that the performance indicators meet system requirements.