Tire Tateral Forces and Sideslip Angle Estimation for Distributed Drive Electric Vehicle Using Noise Adaptive Cubature Kalman Filter

Aiming at the problem that the traditional cubature Kalman filter algorithm needs to know the statistical characteristics of noise first when estimating the critical state of the vehicle, a noise adaptive cubature Kalman filter (NACKF) algorithm is proposed to estimate vehicle critical states. Based on suboptimal and unbiased maximum posterior estimator, the covariance of measured noise is updated in real time and embeds it into the standard Cubature Kalman algorithm to realize the adaptive cubature Kalman filter. Aiming at the influence of coupling characteristics among different subsystems of vehicles on filtering accuracy, a dual adaptive cubature Kalman filter is constructed to estimate the lateral force and the sideslip angle respectively. The distributed modular structure weakens the influence of the coupling characteristics of the system on the estimation accuracy and realizes the real-time accurate estimation of the tire lateral force and the sideslip angle. Finally, to verify the effectiveness of the proposed algorithm, the simulation test and the real vehicle test are implemented. The results show that the estimation algorithm based on dual adaptive cubature Kalman filter has higher estimation accuracy than that based on standard cubature Kalman filter, and it better improves the nonlinear filtering accuracy reduction problem of traditional cubature Kalman filter when the prior statistical characteristics of noise are unknown. © 2019 Journal of Mechanical Engineering.