Robust Adaptive Control for Micro-Vibration Suppression under Multiple Unknown Narrow-Band Disturbances

Abstract

While the robust adaptive control algorithm based on the Y-K parameterization method can significantly enhance the suppression capability of the feedback control system against multiple unknown and time-varying narrow-band disturbances, it requires precise secondary path models for the design of central robust controller. However, in many control systems, obtaining accurate mathematical models is a formidable challenge. As a result, the effectiveness of feedback control systems can be severely compromised or rendered ineffective. This paper presents a direct feedback robust adaptive micro-vibration control algorithm when an accurate secondary path model cannot be obtained. The algorithm is based on the internal model principle and the Y-K parameterization method. It effectively tackles the challenge of parameter design for a central robust controller when the secondary path model is unknown. In addition, we introduce a novel variable step size least mean square (VSSLMS) method, which is demonstrated to offer advantages through system identification compared to other VSSLMS methods. Furthermore, the real-time micro-vibration control experiments are carried out with an active micro-vibration control system. The performance of filtered-X least mean square (LMS), the robust adaptive control algorithm based on LMS, and the robust adaptive control algorithm based on VSSLMS are evaluated through experimental comparisons. The experimental results clearly show that the proposed feedback robust adaptive control algorithm outperforms other control algorithms, especially in scenarios involving dual-frequency sinusoidal disturbances and sudden changes in their spectra and amplitudes.