Enhanced trajectory tracking for autonomous navigation of wheeled mobile robots using an adaptive fuzzy PID controller

With the advancement of autonomous technologies, the need for robust control strategies in unstructured environments is becoming increasingly important. The ability to track a trajectory and accurately control the motion of a robot is a key aspect of mobile robotics and is essential if wheeled mobile robots (WMRs) are to successfully perform tasks and function in the real world. Due to the complex and unstructured working environments, the control systems of WMRs must deal with difficulties such as extreme non-linear behaviors of the dynamic systems, unmodeled parameters of the systems, and external disturbances. This paper presents an adaptive fuzzy gain scheduling PID controller that addresses the challenges posed by structured uncertainties like kinematic wheel slips, random actuator noise, and external disturbances. The controller is based on a robust cascaded control structure that can be used for tracking the desired trajectory. The robustness of the controller is ensured by uniformly ultimately bounded analysis. The control system incorporates adaptive fuzzy logic and PID control to achieve a more advanced level of trajectory-tracking control. The control systems enhance fuzzy logic and PID control with an adaptive capability aimed at improving the systems' robustness against external disturbances. Changes in the robot dynamics or the environment may happen, but the controller can adjust its parameters in real time and achieve the desired result due to the controller's adaptation mechanism. The efficacy of the controller is validated using extensive simulations for tracking complex lemniscate (∞) curves. The results are compared with conventional PID and adaptive dynamic control methods to demonstrate that the proposed controller reduces the RMS tracking error. The results clearly demonstrate that the controller can reject disturbances while achieving precise navigation even for 100% variations in the parameters. The results presented in this paper provide a computationally efficient framework to bridge the gap between kinematics and dynamics for real-time robotic applications.