Comparative analysis of fuzzy logic and neural network based controllers for line tracking task in mobile robots

Abstract

In this paper, design of fuzzy logic and neural network based controllers for line tracing operation of a differentially steered mobile robot is presented. The proposed controllers form a two input, two output systems. Each controller accepts inputs from an array of sixteen infrared sensors grouped as left and right sensor with eight sensors in each group, performs the control algorithm and generates speed commands for left and right motors. The controllers are finally tested and compared on various paths having varying degree of turns. Results have shown the validity of controllers for line tracing task.