Abstract:
Robotics-based solutions often rely on motors, and the precision and accuracy of these robots depend on the quality of their motion control. The most common type of motor is the DC motor, and two approaches for controlling DC motors are available: feedback and feedforward control. Inversion-based feedforward control, the most straightforward approach, uses the system model to provide input. However, this approach is not suitable when system parameters change over time. Feedback control, on the other hand, does not rely on the system model and provides control input based on the error signal. However, feedback control performs poorly in the presence of noise in the feedback signal. In this study, we propose a novel approach combining feedback and feedforward control for DC motor control. We use a smooth switching law that selects the contribution of feedback and feedforward control based on the variance of the error and estimation gain for time-varying systems. Extensive simulation results demonstrate that our approach of smooth switching between feedforward and feedback control provides superior performance in tracing, compared to LQR, in the presence of noisy feedback.