Virtual navigation of mobile robot in V-REP using hybrid ANFIS-PSO controller

Abstract

Abstract: Trajectory following and obstacle avoidance are the main issues for autonomous navigation for mobile robots.

This article describes the virtual navigation of a Pioneer P3-DX differential drive mobile robot which contains two left and right wheels driven by DC motors and a freewheel to ensure its stability in two environments designed on the V-REP virtual platform. At first, we realized navigation in an environment without obstacle by searching for a target by the ANFIS algorithm then by the hybrid algorithm ANFIS-PSO written in Matlab code, and then we realized navigation in a cluttered environment that allows searching a target by avoiding obstacles with the ANFIS algorithm then with ANFIS-PSO.

The objective of PSO algorithm is the minimization of the objective function which is the root mean square error between the actual and the predicted values ??of the ANFIS controller commands which are the left and right velocity of the mobile robot, to give optimum velocity values that will be applied to the robot for better trajectory following and obstacle avoidance.

The Virtual Robot Experimentation Platform software V-REP was used to design our study and simulate the virtual reality of navigation and to work in synchronous mode with the codes created in MATLAB.

A comparative study between navigation with ANFIS and navigation with ANFIS-PSO clearly shows the effectiveness of the PSO which ensures a good response to the system within these physical limitations (velocities, torques) and ensures optimal tracking for the virtual navigation of the mobile robot.

Looking at the experimental results, it is observed that the proposed method provides a more efficient and optimal solution to the problem of target search and obstacle avoidance.

DOI: 10.61416/ceai.v26i1.8797