Multi-secondary Active Noise-proof Window Based On Improved CFxLMS Algorithm

The current research on active noise-proof window is limited to single secondary sound source control system,and the noise reduction range and noise reduction effect are limited.In this paper,the multi-secondary source active noise reduction system is applied to the noise-proof window.The indoor and outdoor space are not comprehensively considered in the existing active noise-proof window model.Besides,the performance of the active noise reduction system is greatly affected by the parameter selection.In the meantime,the lack of noise reduction analysis of real noise is also a common issue.This article has made corresponding improvements to the above problems.Firstly,a multi-secondary active noise-proof window model based on the simplified outdoor sound field coupling indoor sound field is established.Based on the COMSOL software,the simulation calculation is carried out to discuss when there are different numbers of secondary sound sources distributed on the sound insulation window.The distribution of sound pressure in the room.And the number of secondary sources required to achieve standard noise reduction requirements for noise-proof windows of different sizes.Secondly,the multi-secondary active noise-proof window control algorithm is improved and simulated,and the single-channel traditional FxLMS algorithm is extended to the multi-channel control system.In order to solve the problem that the fixed convergence step size is difficult to balance the stability and convergence speed of the active noise reduction system,and the classical variable step function has poor adaptability to the aperiodic signal.An improved variable step size CFxLMS adaptive filtering algorithm is adopted.Using simulation to determine the parameters of the variable step size formula,then extends the algorithm to the multi-channel active control system.By using Matlab simulation,the mixed random signal and the real noise signal are taken as inputs to compare the superiority of the dual-channel CFxLMS algorithm with respect to the traditional FxLMS algorithm.So the effectiveness of improved variable step size algorithms and parameters has been verifiedFinally,the hardware structure design and software programming of the dual-secondary active noise reduction system controller were completed.A suitable experimental environment was chosen to build a multi-channel active noise-proof window experimental platform.The comparison experiment between two-channel active noise reduction and single-channel active noise reduction is carried out to verify the effectiveness of the method.