Improved Trajectory Tracking for Nonholonomic Mobile Robots Via Dynamic Weight Adjustment in Type-2 Fuzzy Model Predictive Control

This paper presents an advanced methodology for trajectory control of non-holographic mobile robots. It addresses the challenges of dynamic environments and system uncertainty by proposing a fuzzy model predictive control (FMPC) system that combines Type-2 fuzzy logic (F2MPC) with model predictive control (MPC) to enhance tracking accuracy and adaptability. A Takagi-Sugeno (T-S) fuzzy model changes the MPC weighting matrices in real-time based on speed and distance errors, while the Type-2 fuzzy system handles uncertainties better than Type-1 systems. Tests using circular and wavy trajectories show that the Type-2 Fuzzy MPC (F2MPC) works better than traditional methods, achieving fewer tracking errors (Integral Squared Error of 0.0011), faster convergence (in 1.2 seconds), and using 65% less energy for movement than conventional MPC. Robustness tests show the controller's stability under disturbances, with minimal deviation and quick recovery. The results highlight the F2MPC's precision, efficiency, and adaptability, making it a promising solution for complex robotic navigation tasks. The study found that Type-2 fuzzy logic and predictive control improve trajectory tracking, paving the path for real-world applications and computational optimisations.