Linear Kalman Filtering with Persistence Predictive Model for Vehicle Slip Angle Estimation Based on Inverse Trigonometric Function

Abstract

Slip angle estimation is extensively employed in mobile robotics. Specifically, the estimation method predicated on the acceleration sensor can compute the slip angle without the necessity for a dynamical model and integration calculations. However, wave distortion of the estimated slip angle emerges due to the inclusion of the inverse trigonometric function in the estimation computation. In this study, we utilized a linear Kalman filter predicated on a persistent predictive model for slip angle estimation. The proposed method modulates the Kalman gain according to the Mahalanobis distance, thereby eliminating the wave distortion of the estimated slip angle. The efficacy of the proposed method is corroborated by a nonlinear vehicle simulation. © 2023 Institute of Electrical Engineer of Japan and Wiley Periodicals LLC.