Fuzzy Observer for Nonlinear Vehicle System Roll Behavior with Coupled Lateral and Vertical Dynamics 2018-01-0559

The study of vehicle state estimation performance especially on the aspect of observer-based control for improving vehicle ride comfort and road handling is a challenging task for vehicle industry. Since vehicle roll behavior with various road excitations act an important part of driving safety, how to accurately obtain vehicle state under various driving scenes are of great concern. However, previous researches seldom consider coupling relation between vehicle vertical and lateral response with steering input under various road excitation. To address this issue, comprehension analyses on vehicle roll state estimation with coupled input are present in this paper. A full-car nonlinear Takagi-Sugeno (T-S) fuzzy model is first created to describe vehicle lateral and vertical coupling dynamics. Then, a T-S model-based fuzzy observer design approach is used to estimate the vehicle state considering nonlinear coupling dynamics of tire lateral force and semi-active suspension actuator force under road excitation input. The stability conditions for used observer are induced using linear matrix inequalities (LMI). The obtained observer is simulated by solving the proposed LMI. Finally, an E-Class (SUV) level car model in a CarSim® software is used to validate the proposed approach under various International Standardization Organization (ISO) road excitation inputs, and simulations results show better estimation accuracy compared to validation data. The research achievements can be further used for improving state estimation accuracy of vehicle system under various road excitation and steering input and provide a more accurate algorithm for vehicle with controllable suspension components in critical situations.