Online Estimation Of Vehicle Power Battery Power Stat

With the shortage of energy such as oil and the popularization of the concept of new energy,the development of new energy vehicles has become an important measure to solve the energy crisis.The battery management system is responsible for ensuring that the battery is in a safe,stable and efficient working state.Battery state estimation is one of the main functions of the battery management system.The battery state estimation includes the battery’s state of charge(SOC),the battery’s power state(SOP),and the battery’s state of health(SOH),etc.where SOP represents the charge and discharge capacity of the battery in the current state.Accurate SOP estimation is important for efficient battery utilization.It is indispensable for the high-precision,high-efficiency and reliable management of battery management system and even new energy electric vehicles.In the SOP estimation method,the equivalent circuit model(ECM)is usually used to simulate the dynamic characteristics of the battery,and then the SOP values under different durations are calculated by using the parameters of the battery equivalent circuit model.It can be considered that the accuracy of battery equivalent circuit model determines the accuracy of battery SOP estimation.However,there is always a contradiction between the complexity and accuracy of the model.A simple model usually can not accurately reflect all the dynamic characteristics of the battery,which may bring error identification to the parameters.The complex battery equivalent circuit model can not be applied to the on-line and real-time calculation of electric vehicles due to the power exponential rise of the amount of calculation.In order to solve this problem,this paper first proposes a new equivalent circuit model,that is,the grey box model of vehicle power battery equivalent circuit based on auto regression(AR)model.The AR model with linear structure and relatively simple calculation is used to fit the polarization voltage drop of power battery.Its advantage is that it can be updated online and has good dynamic characteristics.It can effectively improve the accuracy of the model.Through simulation verification and comparison,the accuracy of RC-ECM is higher than that of RC-ECM.Aiming at the problem that some parameters are inaccurate in the process of parameter identification,extended Kalman filter(EKF)and recursive least squares(RLS)are combined for parameter identification.According to the time-varying characteristics of battery equivalent circuit model parameters,a two-scale parameter identification algorithm is proposed by combining extended Kalman filter with recursive least squares.While ensuring the accuracy of the model,the computational complexity of the algorithm is reduced.Based on the above AR-ECM model,this paper proposes to realize long-term and short-term online estimation of electric vehicle power battery SOP under the restriction of multiple factors,that is,comprehensive consideration of SOC limitation,factory setting limitation,and battery charge-discharge cut-off voltage.Finally,the open dataset of CALCE battery experiment at the University of Maryland is selected to verify the effectiveness and applicability of the proposed multi state joint estimator SOP algorithm based on AR equivalent circuit model under dynamic stress test(DST).It has good robustness and verifies the practicability of this algorithm.