Research and design of intelligent estimation method of EV Vs based on multiple Kalman filters

Vehicle dynamics simulation technology is widely used in the development process of intelligent vehicle control algorithm. On the one hand, vehicle dynamics simulation can be applied to the initial verification of the control algorithm, and test the effectiveness of the algorithm before the real vehicle experiment to save the time and cost; on the other hand, vehicle dynamics simulation is also used for the optimization design of the control algorithm parameters. Researchers use the simulation technology to construct the controller evaluation training environment and optimize the controller parameters repeatedly^[1],[2]. It is a key problem to improve the accuracy of the state estimation algorithm. The model-based state estimation methods mainly include Kalman filter, extended Kalman filter, traceless Kalman filter and volume Kalman filter^[3],[6]. In this paper, we establish a vehicle dynamics simulation model based on the above mechanism, analyze the advantages and disadvantages of various Kalman filtering algorithms, and verify the algorithm, which improves the accuracy and intelligence of electric vehicle state estimation in the future. According to the analysis of the experiment results, the accuracy of the trace Kalman filter algorithm and the volume Kalman algorithm based on the seven DoF model is higher than that of the extended Kalman filter, and the volume Kalman filter algorithm and the estimation performance of the pendulum angle velocity and center of mass offset angle, which is improved compared with the extended Kalman filter performance.