Research And Implementation Of On-line SOC Estimation For Vehicular Lead-Acid Battery

Lead-acid battery takes the role of engine start and auxiliary power supply in the vehicles. It is an important component of the whole vehicle power system. The state of charge (SOC) of lead-acid battery is an important parameter of battery inner state. SOC, one of the inputs of power management strategy in energy-saving and environment-friendly vehicles, is directly related to the life of battery and the effectiveness of vehicle energy management.This thesis is supported by National Natural Science Foundation and enterprise cooperation project. The algorithm of online SOC estimation for vehicular lead-acid battery and its implementation in battery sensor are researched. Firstly, the simplified Randles equivalent circuit model that conforms to lead-acid battery characteristics is established through the analysis on the existing battery models. The covariance matrix of convergence algorithm combined with Kalman filter is used to online identification of model parameters. Secondly, the improved Kalman filter is respectively combined with extended Kalman filter and unscented Kalman filter to estimate the SOC of lead-acid battery, and the accuracy of two algorithms is compared. The influencing factors of SOC estimation accuracy are analyzed. Then, the above method of SOC estimation is implemented on MC9S12ZVCA128 main chip. The prototype node of battery sensor is developed. The circuits of parameter acquisition and LIN communication are designed. The program of parameter acquisition, SOC estimation and LIN communication are developed. Finally, the simulation model based on Simulink and S function is established. The battery test bench is set up. The PC test software is designed. The discharging experimental scheme is proposed as well as the specific experimental procedures and data gathering method are developed. The results of battery discharging experiments show that the algorithm of parameter identification and SOC estimation can effectively improve the accuracy of SOC estimation, and the parameter detection and SOC estimation of the battery sensor are accurate and reliable. This study is of great significance for strengthening the protection of the battery, improving the effectiveness of power management and exploring the implementation of high-precision complex algorithm in battery sensor.