A federated filter design of electronic stability control for electric-wheel vehicle

Abstract

The aim of this study is to improve the convergence speed and accuracy of the filter in the ESC (Electronic Stability Control) of electric-wheel vehicle. A federated filter is designed based on an improved UKF (Unscented Kalman Filter). The federated filter is composed of a vehicle speed filter and a road adhesion coefficient filter. The two filters are connection and correction each other. In the improved UKF, a scaled minimal skew sampling strategy is used to reduce the number of sampling points and avoid the local effects. A tracking adjustment factor and a resisted demission error factor are added to the improved UKF algorithm. These factors are used to enhance the tracking performance and eliminate the outliers of the system measured value. An electric-wheel vehicle dynamics model is established for the federated filter. Simulation experiments of the vehicle speed estimation and the road adhesion coefficient estimation were done. The road adhesion coefficient was 0.8 and 0.2. The initial vehicle speed was 100 and 90 kilometers per hour. The results shown the federated filter could shorten the delay time and reduce the overshoot. The federated filter is fit for the ESC of electric-wheel vehicle.