A Cognitive System Design for Mobile Robot Based on an Intelligent Algorithm

Abstract

This paper presents a cognitive system based on a nonlinear Multi-Input Multi- Output (MIMO) Proportion Integral Derivative (PID) Modified Elman Neural Network (MENN) controller and the Square Road Map (SRM) method to guide the mobile robot during the continuous path-tracking with collision-free navigation through static obstacles. The proposed cognitive system consists of two parts: the first part is to plan the desired path for the mobile robot with the static obstacle environment in order to determine the target point and to avoid the obstacles based on the proposed square road map algorithm. The second part is to guide and track the wheeled mobile robot on the desired path equation based on the proposed nonlinear MIMO-PID-MENN controller with the intelligent algorithm. The Particle Swarm Optimization (PSO) is used to on-line tune the variable control parameters of the proposed controller to get the optimal torques actions for the mobile robot platform. Based on using the MATLAB package (2017), the numerical simulation results show that the proposed cognitive system has high accuracy for planning the desired path equation in terms of avoiding the static obstacles with smooth and short distance and generating a perfect torque action of (0.7 N.m) without a saturation state of (3.07 N.m), which leads to minimize the tracking pose error for the mobile robot to the zero value approximation. These results were confirmed by a comparative study with different nonlinear PID controller types in terms of number of iterations and the performance index.