A novel adaptive step-size hybrid active noise control system

Abstract

The conventional hybrid active noise control (HANC) system adopts the standard filtered-x least mean square (FxLMS) algorithm to adapt its control filters. However, the constant step size in the algorithm leads to not only the contradiction between convergence rate and steady-state noise reduction, but inadequate robustness under certain noise circumstances. For performance improvement, a new adaptive step-size method is incorporated into the HANC (ASHANC) system to optimize the adaptation of the filters. It combines the strengths of the variable step-size FxLMS (VSFxLMS) algorithm and the modified normalized FxLMS (MNFxLMS) algorithm. Different step-size adjustment strategies can be obtained by setting specific parameters. Furthermore, to obtain the best performance of the system, we develop a method to determine the involved key parameters on the basis of the improved particle swarm optimization (IPSO) algorithm. The objective function of the IPSO algorithm takes the convergence rate and noise reduction into consideration to search for optimum values. Simulation results demonstrate that the proposed ASHANC system achieves faster convergence rate, better noise reduction, and enhanced robustness than the other investigated systems.