Research On On-line SOC Estimation Of Vehicular Lead Acid Battery Based On Battery Model

Lead-acid battery used for engine starting and auxiliary power supply in the car is an important component of the vehicle power system. Lead acid battery state of charge (SOC) is an important parameter of battery inner state, which is closely related to some nonlinear factors such as battery temperature, state of health(SOH) and other initial battery state. How to estimate the lead-acid battery SOC in real time is directly related to the effectiveness of the SOH and vehicle energy management.Firstly, the chemistry working principles of lead-acid battery and some common performance parameters are introduced simply. Through experimental method of battery voltage, capacitance, resistance and other characters related to SOC, battery charge and discharge characteristics are studied.Secondly, according to the relevant characteristics of the battery, combined with lead-acid batteries Randles equivalent circuit model, a RC equivalent circuit model which reflects discharge voltage hysteresis effects and ignores the self-discharge is proposed. In order to ensure system accuracy, the battery model the dynamic parameters are estimated online. Thirdly, in view of the circuit model, the state equation and linear output equation combined with the open-circuit voltage of the lead-acid battery SOC relationship characteristics is established. By real-time adjusting covariance optimization iteration steps, kalman filtering algorithm is improved. Joint with the time-varying model parameters identification into dynamic system equations, a method for SOC estimation online of Dual Extended Kalman filter (DEKF) is proposed. Finally, a discharge experimental platform is designed for spiral wound VRLA battery. With the experimental data collected, AH integration method and fixed parameters Kalman filtering method are compared to the proposed algorithm. The final results demonstrate the effectiveness of the method proposed, which does not require batteries offline estimate model parameters, and it is easy to get more accurate battery model parameters and the SOC. Compared to traditional AH integration method and Kalman filter method, the DEKF can ensure SOC estimation accuracy and is more suitable for online SOC estimation.