Output Feedback Predictive Steering Control in Autonomous Driving : An Interval Type-2 T-S Fuzzy-Model-Based Approach

In this paper, the problem of output feedback predictive steering control for path tracking in autonomous vehicles is studied based on interval type-2 fuzzy-model-based approach. Due to the complex road conditions and different maneuvers, the vehicle experiences fluctuations in velocity, and furthermore, certain physical characteristics of the vehicle are not accessible for control. The suggested approach involves the utilization of fuzzy modeling techniques to develop an interval type-2 T-S fuzzy model that captures the vehicle’s lateral dynamics, taking into account the longitudinal velocities velocity change. A Longberg observer is developed for the state estimate and the corresponding conditions for the estimation error to be confined within a minimal ellipsoidal robust positively invariant set are proposed, which is involved in an H∞-type cost function of predictive control to further improve the transient output feedback control performance. Then, an interval type-2 fuzzy-model-based predictive controller is established by addressing a Min-max optimization problem. Finally, in order to demonstrate the effectiveness of proposed methods, some validation results with Carsim/MATLAB joint simulation are provided.