Running States Estimation of Autonomous Four-Wheel Independent Drive Electric Vehicle by Virtual Longitudinal Force Sensors

Abstract

Exact sideslip angle estimation is significant to the dynamics control of four-wheel independent drive electric vehicles. It is costly and difficult-to-popularize to equip vehicular sensors for real-time sideslip angle measurement; therefore, the reliable sideslip angle estimation method is investigated in this paper. The electric driving wheel model is proposed and applied to the longitudinal force estimation. Considering that electric driving wheel model is a nonlinear model with unknown input, an unknown input estimation method is proposed to facilitate the longitudinal force observer design, in which the adaptive high-order sliding mode observer is designed to achieve the state estimation, the analytic formula of longitudinal force is obtained by decoupling electric driving wheel model, and the longitudinal force estimator is designed by recurrence estimation method. With the designed virtual longitudinal force sensor, an adaptive attenuated Kalman filtering is proposed to estimate the vehicle running state, in which the time-varying attenuation factor is applied to weaken the past data to the current filter and the covariance of process noise and measurement noise can be adjusted adaptively. Finally, simulations and experiments are conducted and the effectiveness of proposed estimation method is validated.