Iterative Learning Disturbance Estimation-Based Fuzzy Control System of Four-Wheel-Steering AGVs

This paper investigates a strongly nonlinear system with coupled time-varying parameters related to vehicle speed and adhesion coefficient, as well as repetitive disturbances under practical industrial operating conditions. A fuzzy tracking control system framework is proposed, which consists of a Takagi–Sugeno (T-S) fuzzy observer with proportional-derivative (PD)-type iterative learning disturbance estimation and a T-S fuzzy P controller with disturbance compensation. Using the sector nonlinearity approach with properly selected premise variables, an exact T-S fuzzy representation of four-wheel-steering (4WS) vehicle dynamics is constructed. The T-S fuzzy observer is designed based on a common Lyapunov function and the linear matrix inequality (LMI) technique. Furthermore, an iterative learning law is developed to suppress repetitive disturbances across successive iterations. By employing a T-S fuzzy proportional controller together with disturbance compensation, the proposed scheme realizes high-precision path tracking. Joint time- and iteration-domain Lyapunov analysis derives LMI-based co-design conditions for all control gains, ensuring that tracking and observation errors converge to bounded regions. Simulation results demonstrate fast error convergence and satisfactory tracking performance of the proposed method under various operating conditions.