Design Of Soc Estimation Algorithm For Power Supply Battery Based On Kalman Filter

The accurate estimation of the state of charge for battery is an important basis for the control of charge and discharge of the electric vehicle battery and the optimal management of power. The working life of battery and the performance of car are directly influenced by the accuracy of the estimation of the state of charge for battery. The research of the estimation of the state of charge for power supply battery of electric vehicle is mainly composed of the following parts:1. The mapping approximation is used to achieve the linearization of model. The ambient temperature scale factor and charge-discharge rate scale factor are introduced to determine the equivalent Coulombic efficiency, and the Kalman filtering algorithm based on composite model is designed. The simulation results show that the new algorithm has good ability to correct the accumulation error and initial error.2. The weighted statistical linear regression method is used to achieve the linearization of model function. For the system characteristic of linear state equation of composite model of battery, the combination of the standard Kalman filtering algorithm and the Kalman filtering algorithm based on weighted statistical linear regression method is used. With the introduction of the singular value decomposition, the Kalman filtering algorithm based on singular value decomposition is designed. The simulation results show that the new algorithm has better computational efficiency, and better convergence rate and simulation accuracy than the Kalman filtering algorithm based on composite model.3. The suboptimal fading factor based on the principle of strong tracing is introduced, the Kalman filtering algorithm based on singular value decomposition is designed, and the strong tracing ability to deal with the forced condition and the robustness to deal with the inaccurate of the model are achieved. The simulation results show that the new algorithm has higher simulation accuracy and faster convergence rate than the Kalman filtering algorithm based on composite model and the Kalman filtering algorithm based on singular value decomposition.