Kinematic control of an dc motor by a comparative approach between a digital PID controller and two soft computing controllers (fuzzy and neuro-fuzzy)

Abstract

The rise of mobile line-following robots in various industrial and civil fields of activity suggests the advent of increasingly robust autonomous mobile robots. However, this development is still marked by several major challenges such as navigation with obstacle avoidance in unknown or known environments and, more particularly, the fluidity of their mobility during trajectory turns due to their kinematics. The objective of this paper is to provide one of the appropriate solutions to this problem. Our contribution will focus on the estimation of the model of our DC motor via data from the acquisition of speed measurements of the latter during the experiment carried out under Arduino and Matlab, then on the design of speed controllers for the control of the DC motor of a mobile car-type robot and finally on a comparison of the responses between the classic PID controllers and those of Soft Computing (fuzzy controller and fuzzy neuro). The implementation is carried out with an Arduino microcontroller for the control of the DC motor (FIT0521) via a L298N double H-bridge motor driver, and on the other hand by implementing the different controllers mentioned above in the Matlab/Simulink environment. The simulation results of the three control models showed the robustness of the smart controllers compared to the classical one for non-linear systems. These results were confirmed experimentally on our DC motor "FIT0521" used for the circumstance in place of the motors of the manufacturer "Elegoo" in the mobile robots of car type and the analysis of the curves thus obtained.