Velocity control of a unicycle type of mobile robot using optimal pid controller

Abstract

A unicycle model of control a mobile robot is a simplified modeling approach modified from the differential drive mobile robots. Instead of controlling the right speed, VR and the left speed, VL of the drive systems, the unicycle model is using u and ω as the controller parameters. Tracking is much easier in this model. In this paper, the dynamic of the robot parameter is controlled using two blocks of Proportional-Integral-Derivative(PID) controllers. The gains of the PID are firstly determined using particle swarm optimization(PSO) in offline mode. After the optimal gain is determined, the tracking of the robot’s trajectory is performed online with optimal PID controller. The achieved results of the proposed scheme are compared with those of dynamic model optimized with genetic algorithm(GA) and manually tuned PID controller gains. In the algorithm, the control parameters are computed by minimizing the fitness function defined by using the integral absolute error(IAE) performance index. The simulation results obtained reveal advantages of the proposed PSO-PID dynamic controller for trajectory tracking of a unicycle type of mobile robot. A MATLAB-Simulink program is used to simulate the designed system and the results are graphically plotted. In addition, numerical simulations using 8-shape as a reference trajectory with several numbers of iterations are reported to show the validity of the proposed scheme.