Design Of Indoor Active Noise Reduction System In Office Area

Noise pollution in the living environment has become one of the most important factors affecting people’s quality of life.Traditional passive noise control can only remove most of the high-frequency noise,but the low-frequency noise,which is more harmful to human body,is not effectively eliminated.Active noise control can make up for the defects of passive noise control.In order to effectively reduce the impact of indoor low-frequency noise in office areas,this paper takes the prevalent indoor air conditioning operation noise as the object and carries out the research of indoor active noise reduction system in office areas,and the main work completed in this paper is as follows.(1)Based on the research of active noise reduction technology,the structure of indoor active noise reduction system is determined.Through simulation analysis,the convergence performance of the LMS algorithm is obtained to be good,and the noise reduction performance can be maintained in the use scenario with small SNR,which verifies the effectiveness of the LMS algorithm for active noise control and applies it to the indoor active noise reduction system in this paper.(2)By measuring and analyzing the actual operation noise of indoor air conditioners,the lowfrequency and narrow-band characteristics of air conditioner noise are obtained,and a singlechannel feed-forward narrow-band active noise reduction system is determined according to this characteristic.The existence of secondary paths in practical applications can cause time delay problems,resulting in a weakened noise reduction effect of the system,which leads to the introduction of the Fx LMS algorithm.The offline modeling method and online modeling method of secondary path modeling are compared,and the online modeling method of secondary path is selected by combining the characteristics of indoor noise.It is verified through simulation that the online modeling method has higher modeling accuracy and can effectively eliminate the influence of time delay problem.The iteration step size and filter length,which have a great influence on the performance of the noise reduction algorithm,are investigated,and it is concluded through simulation that a larger iteration step size and filter length will speed up the convergence speed of the algorithm but increase the mean square error of the algorithm under the premise of ensuring the stability and non-divergence of the algorithm.In the algorithm design of indoor active noise reduction system,it is necessary to weigh the mean square error and convergence speed,and choose the appropriate iteration step and filter length.(3)To address the problem that the fixed iteration step size in the Fx LMS algorithm affects the stability of the system,an improved variable step size Fx LMS algorithm is proposed,which realizes the adaptive adjustment of the system by synchronizing the change of the iteration step size of the primary control filter and the secondary path modeling filter to enhance the stability of the algorithm.Meanwhile,to address the problem of modeling noise affecting the noise reduction effect in the Fx LMS algorithm,a strategy of monitoring the input signal power of the algorithm is used to control the injection timing of modeling noise in order to reduce the impact of modeling noise on the noise reduction effect of the system.The simulation comparison with the existing algorithm proposed by Akhtar shows that the improved algorithm has higher secondary path modeling accuracy,better steady-state performance and faster convergence speed,and can achieve a maximum noise reduction of 12.2 d B.(4)Based on the improved variable step size Fx LMS algorithm,the hardware platform of the indoor active noise reduction system was built.After joint simulation tests,the comparison test results show that: the residual error results of the actual test of the active noise reduction system basically match with the residual error results obtained from the Matlab simulation of the noise reduction algorithm;the actual secondary path weight coefficients of the active noise reduction system basically match with the secondary path weight coefficients estimated by the modeling of the noise reduction algorithm.The feasibility of the active noise reduction system designed in this paper is verified.