Effect of filtering techniques on the derivative term in fuzzy logic controller for dc motor position control

Abstract

Direct Current (DC) motors are fundamental components in various industrial and automation systems, valued for their precision and controllability. Traditional control methods, such as Proportional-Integral-Derivative (PID) controllers, often require robust mathematical models and are susceptible to performance degradation under non-ideal conditions. This study investigates the implementation of Fuzzy Logic Controllers (FLC) for real-time DC motor position control, with a focus on analyzing the impact of different derivative approaches. To construct a comprehensive mathematical model of the DC motor system, both white-box and black-box system identification approaches were employed. The white-box method utilized physical principles of the motor, while the black-box method relied on empirical input-output data. The Transfer Function-Based Derivative (TFD) and Second-Order Filtered Derivative (SOFD) techniques are evaluated for their maintaining system responsiveness. A test setup utilizing an STM32F4 discovery kit was developed, and the performance of both derivative approaches was compared using a repeating stair sequence as the reference input. The experimental results showed that both techniques performed successfully, but the SOFD method demonstrated a more effective error reduction. The findings offer insights into derivative filtering techniques, highlighting the benefits of incorporating advanced filtering strategies in FLC-based control systems.