A hybrid-like observer of body slip angle for electric vehicle stability control: Fuzzy logic and Kalman filter approach

A hybrid-like observer for vehicle body slip angle (beta) is proposed, in which a local approximation of the nonlinear tire model is adopted. Local equivalent vehicle models for observer design are built by linear approximation of vehicle dynamics respectively for low and high lateral acceleration operating regimes. Fuzzy logic approach is adopted to combine the local observer models so as to deal with the nonlinear nature of vehicle dynamics. The local observers are designed as linear observers with Kalman filter theory to overcome the influence of system noise. The derivation of this hybrid-like observerpsilas state equations and the estimation mechanism is developed. The fuzzy rules can establish the qualitative relationships among the variables concerning the nonlinear and uncertain nature of vehicle dynamics, such as the saturation of tire force and the influence of road adherence. By adaptation mechanism designing of membership functions in the fuzzy rules, the quantitative accuracy and adaptive performance of the system can be satisfied, which is verified by simulations and experiments.