ACQUISITION OF A STATE-SPACE MODEL FOR ONE-DIMENSIONAL ACTIVE NOISE CONTROL

摘要

The acquisition of a state-space model for Active Noise Control (ANC) for a one-dimensional acoustic duct system is discussed. A complex-valued state-space model with infinite dimension is derived theoretically. The key system coefficient-acoustic terminal impedance is determined based on system identification method. By checking the eigenvalue' s consistency, the acoustic terminal impedance is determined by taking average of its estimations corresponding to each resonance frequency. The proposed approach is further extended into a peak-by-peak identification approach so as to reduce the complexity and computation load caused by potential high-order system identification. Compared with most existing models and approaches for ANC design, the unique feature of the proposed model and method is that the physical meanings of all system coefficients are kept in the developed model, which provides sufficient details for supporting ANC design and analysis.