Piecewise Disturbance-Rejection Control of Systems With Multifrequency Disturbances

This study tackles the problem of rejecting disturbances with different frequencies for a control system subject to measurement noise. The equivalent-input-disturbance (EID) approach is widely employed to address this problem, as it actively estimates and compensates for disturbances. However, its performance in handling multifrequency disturbances, particularly high-frequency ones, is compromised by measurement noise. This work presents a multifrequency EID (MEID) approach characterized by a new multilevel structure, where each level operates at a distinct bandwidth to overcome this limitation. First, the multilevel structure is described in detail. One level is designed to handle constant and low-frequency disturbances using an EID estimator with a low-pass filter. The others are constructed to handle disturbances in different frequency domains using EID estimators with bandpass filters. Then, a comprehensive comparison with two recently proposed alternative structures is provided. A frequency-domain analysis reveals that the MEID approach not only enhances disturbance rejection across a broad spectrum but also improves measurement noise suppression. Thereby, it successfully decouples the conflicting design requirements inherent to conventional methods. Finally, experimental results validate the method’s effectiveness, demonstrating its superior performance over other related approaches.