Neuro-Fuzzy Optimization of Navigation Time for a Differential-Drive Mobile Robot Controlled by a Fuzzy Controller

This paper presents a comparative study of two control strategies: fuzzy logic and adaptive neuro-fuzzy inference system (ANFIS) for autonomous guidance of a differential-drive mobile robot. The robot executes goal-seeking and reactive obstacle-avoidance tasks in a MATLAB Simulink environment using the Mobile Robotics Simulation Toolbox. Initially, a fuzzy logic controller with expert-defined IF–THEN rules generates linear and angular velocity commands while logging state and control data into a training matrix. These data are then used to train an ANFIS model, which is redeployed under identical simulation conditions. Both controllers are compared based on path-tracking accuracy, obstacle-avoidance robustness, and control-loop execution time. Simulation results indicate that the ANFIS controller reproduces the fuzzy logic decision boundaries with reduced computational latency, demonstrating the effectiveness of data-driven neuro-fuzzy models for real-time mobile robot control.