Research On The Estimation Method Of The State-of-charge Of Lithium-ion Batteries For Electric Vehicles

The deterioration of the environment and the tight supply of energy have become more and more serious,and the demand for quality environment has become stronger.Therefore,the new energy electric vehicle,which can reduce the emission of a large number of polluted gases and save energy and protect the environment,has ushered in a historic development opportunity.Among the three main components of the new energy electric vehicle-battery,motor and electronic control,the most important one is the battery which provides the kinetic energy of the vehicle.The development of battery technology directly guides the development trend of electric vehicles.For example,the effective improvement of maximum range capability will provide more development possibilities for pure electric vehicles.The key of battery technology research lies in battery management system,which can realize the real-time monitoring of battery status,complete the safety protection of battery packs and fault warning,and ensure the safe and efficient use of energy in the operation of electric vehicles.In this dissertation,the ternary lithium batteries as the research object,and the core issue of the battery management system,the estimation of the state of charge of the lithium battery,is studied and analyzed.The specific results are as follows:(1)The working principle of lithium battery was studied.The related tests were carried out on the ternary lithium battery,and the capacity characteristics,voltage characteristics,internal resistance characteristics and the influencing factors of state of charge were analyzed.According to the basic characteristics of the lithium battery and the existing model,a dual-polarization(DP)composite model of the lithium battery is established.The model not only takes into account the temperature,charging and discharging direction,charging and discharging rate,but also takes into account the bipolarization effect of lithium batteries.The parameters of the model are identified by on-line least square method.(2)The unscented Kalman algorithm(UKF)is used to estimate the state of charge of lithium batteries.Firstly,the basic principles of the unscented Kalman filter algorithm are analyzed,including the unscented transform and the basic Kalman filter idea.The extended Kalman filter(EKF)is introduced.The UKF and EKF were used to estimate the state of charge,the effectiveness of the two algorithms is verified,and the convergence of UKF is not as good as that of EKF.(3)For the difficulty of calculating the charge status of power lithium battery(e.g.poor estimation and reliability),this dissertation presents the way of Unscented Kalman Particle Filter(UPF)based on the online recursive least square method matched by DP battery model.The results show that,compared to the EKF,Extended Kalman particle filter(EKPF)and the UKF,the online parameter identification assisted by UPF shows stronger robustness effect and convergence stability.Besides,in comparison to the off-line parameter identification with UPF,the estimated accuracy of27.59%was improved by using the online parameter identification.