Analysis and Implementation of a DC Motor System Based on Arduino-Simulink Matlab Using Linear Quadratic Regulator (LQR) Control

Abstract

This study presents the comprehensive process of identifying a DC motor system using experimental techniques facilitated by MATLAB's identification tools. Following the acquisition of a mathematical model for the DC motor system, an optimal control technique, specifically the Linear Quadratic Regulator (LQR), is designed to evaluate the step response of the system and enhance its dynamic performance. In this research, a DC motor identification module integrated with Arduino is developed to simplify and streamline the modeling process of DC motors, adopting first-order and second-order model approximations. This module acts as a bridge between Arduino hardware and MATLAB’s Simulink environment, enabling seamless input-output data acquisition for system identification purposes. The identification process resulted in a DC motor model constructed using a second-order Auto Regressive with Xogenous inputs (ARX) structure, which was subsequently used as the basis for the LQR control design. The LQR control technique was implemented by computing the Q matrix elements through the multiplication of the system's transpose C matrix with its original C matrix, while the R matrix parameter was experimentally tuned to a value of 0.000001 to achieve optimal balance between system performance and control effort. Comparative analysis demonstrated that the LQR control method significantly improved the system's response, achieving a time constant of 0.02 seconds, which outperformed the conventional PID controller in terms of responsiveness and stability.