Intelligent Method of Adaptive Neuro-PWM Speed Control for a Mobile Robot in a Constrained Workspace

Abstract

This paper proposes an intelligent method for adaptive Neuro-PWM speed control of a mobile robot in a constrained workspace, integrating a speed error model, an adaptive neural network controller, PWM signal generation, and a dynamic model of the DC drive. Numerical simulations performed under varying reference speeds, external loads, and obstacle distances confirmed the effectiveness of the proposed method, demonstrating adaptive PWM signal correction and stabilization of the neural network weights, thereby providing a basis for its practical implementation in mobile robot control systems.