A Unified Control Strategy for the Active Reduction of Sound and Vibration

Abstract

The generalized minimum variance (GMV) control strategy is proposed as an effective strategy for active sound and vibration control systems. The GMV strategy is shown to unify well-known adaptive filtering approaches based on LMS-type algorithms and purely feedback strategies as used in other types of control systems. The short comings of the LMS adaptive filtering systems for active sound control are clearly un covered. Pure adaptive filtering emerges from the GMV strategy merely when the extra neous noise at the error sensor is Gaussian noise, the controller complexity is restricted and no fast tracking of system changes is required. The necessary additional identification of the secondary acoustic path is an extra problem of the adaptive filtering approach. Sim ulation experiments with real and synthetic data verify the wide applicability of the GMV control strategy and the limited efficacy of (LMS) adaptive filters for use in active sound reduction systems.