Systems theory and the active control of duct noise

Abstract

The behavior of an acoustical system may be represented using either a transfer function or impulse response. This representation enables an electrical model of the acoustical system to be created. The output of this model may be used to drive an acoustical source and produce a secondary wave to cancel an undesired primary wave. A system identification process generates the required model using an adaptive parameter estimation scheme. A variety of alternatives are available for the model structure as well as the parameter estimation scheme. These include a finite impulse response (FIR) or infinite impulse response (IIR) digital filter with parameters obtained using a least‐mean‐square (LMS) or recursive least‐mean‐square (RLMS) adaptive algorithm, respectively. Since the process is one of adaptive control, the filtered‐X or filtered‐U algorithm can be used to compensate for various transfer functions, identified on an off‐line or on‐line basis, following the adaptive model. Results will be presented from ...