Distributed Electric Vehicle State Parameter Estimation Based on the ASO-SRGHCKF Algorithm
Yaming Liu, Rongyun Zhang, Peicheng Shi, Linfeng Zhao, Feng Yongle, Yufeng Du
Abstract
To accurately obtain the state parameter information of a vehicle, a square root generalized high-order cubature Kalman filter (CKF) estimation algorithm based on the atomic search optimization algorithm (ASO-SRGHCKF) is proposed. On the basis of the high-order CKF, using the generalized cubature rule instead of the cumbersome spherical cubature rule, the algorithm’s weights and cubature points are calculated directly. Then, the square root filtering technology is introduced, and the square root generalized high-order CKF (SRGHCKF) algorithm is derived by replacing the Cholesky decomposition with orthogonal triangle (QR) decomposition. To lessen the estimate error brought on by the noise covariance matrix’s uncertainty, the atomic search optimization (ASO) algorithm is used to optimize it, and the algorithm is utilized for the state parameter estimation of distributed electric vehicles. MATLAB/CarSim cosimulation and experiments evidence that the ASO-SRGHCKF algorithm produces more accurate estimation results and faster convergence than the HCKF algorithm and can precisely obtain the vehicle’s state parameter information.