A Partial-updated adaptive algorithm applied to multi-channel active control system for time-varying secondary paths

Abstract

For a multi-channel active control system with time-varying secondary paths, the more the number of the channels the system has,. the greater the computational load will be increased. In this paper, a new Partial-updated Filtered-x LMS (PFxLMS) Algorithm is proposed to reduce computational load. Firstly a Filtered-x LMS (FxLMS) Algorithm based on on-line modeling for secondary paths is derived. And then according to time-varying characteristics for secondary paths the weight coefficients of adaptive filters for active control and adaptive modeling are respectively updated partially such that the computational load can be largely reduced. The simulation results showed that the proposed algorithm reduces considerable noise with reasonable computational complexity compared with standard FxLMS and runs well.