Active vibration control systems in the frequency and sub-band domain.

Abstract

Active noise and vibration control has been the subject of intense study in the last two decades due to the increased speed in digital signal processors and the technological development and manufacture of smart materials. This dissertation analyzes an active control system using adaptive digital signal processing techniques and applies it to the vibration reduction of hard disk drives (HDD). Specifically, this work focuses on the implementation of the adaptive algorithm in the frequency and sub-band domains for performance improvement.

In this dissertation, selective adaptation in the frequency domain is proposed to alleviate the constructive interference associated with a feedback active control system. A new sub-band adaptive filter architecture without a signal path delay is proposed, and the associated adaptive algorithm is derived. This delayless sub-band algorithm can be applied to the active control systems to improve the convergence rate and trade-off the performance from the various sub-bands. The resulting side effect of the error path delay of the analysis filter bank is analyzed, and two compensation methods are proposed to increase the performance. The frequency domain method and the sub-band decomposition technique are then combined to improve the overall performance. The single-channel active control system is extended to the multiple-channel active control system to reduce the vibration of complex mechanical structure. The optimal performances of three variants of the feedback control system have been derived in terms of the correlation coefficients of the primary disturbances and the impulse responses of the secondary paths. Real time and simulation results are performed to verify the efficiency of the proposed algorithms and techniques.