SOC Estimation Algorithm Base On Lithium-ion Battery

As the environmental problems and the energy crisis increasingly serious, vigorously developing the electric car, and accelerating the transformation of traffic energy is an important way to realize the sustainable development of automobile industry, battery is the power source of the electric car, which is the core component of the working car, accurate estimation of state of charge (SOC) has important influence on safe, reasonable use of the electric car, battery life and the improvement of the vehicle's performance, but a lot of factors impact battery SOC estimation, so it is difficult to ensure the estimation accuracy. Based on the Li-ion batteries, studying battery model and algorithm of SOC.This paper research the battery material for affecting the performance of the battery, describes the Li-ion batteries'internal structure and working principle, implementing battery characteristics experiments,which confirmed that temperature, charge /discharge rate, SOC have an important effect on OCV, resistance, and charge/discharge efficiency. On this basis, combining with the analysis of the equivalent circuit model, puts forward a simplified GNL model, by using the OCV (open circuit voltage)-SOC curve to estimate model parameters of the initial value, and using the recursive least square (RLS) method to approach the actual parameters value.This paper briefly expounds the commonly methods of SOC estimation, their advantages and disadvantages, based on the extended Kalman filter (EKF) algorithm's analysis, modified EKF algorithm, combined with simplified GNL equivalent circuit model to list the space state equation, estimated the SOC, at the same time, modified Ah method combine with EKF algorithm improves the estimation accuracy and the anti-interference performance.Using the MATLAB to simulate, contrast the results and the output of the BMS validated experiment platform. HPPC cycle experiment shows that the simplified GNL model not only improves the estimate time, making the recursive least square method to update parameters in real-time, but without great influence on model precision, it accurately can reflect the battery characteristics; By comparing with modified EKF algorithm and Ah method, verified the proposed algorithm enhances the precision of SOC estimation.