Theoretical And Experimental Study On Active Noise Control In 3D Space

Noise pollution has seriously affected people's production and life,and it is imperative to control noise pollution.At present,there are a variety of active noise control systems,but there is no active control system for noise reduction in the local space near the headrest of the seat.This topic focuses on the active control of low-frequency noise near the headrest,providing a stable noise-reduction environment for the three-dimensional local space near the head of the passenger.In the active control system experiment,the primary noise emitted by the primary sound source is collected by the reference microphone as a reference noise signal.After the reference noise signal is processed by the controller,the secondary sound source is driven to form a canceling sound wave in the noise reduction region.The canceling sound wave is coherently superimposed with the desired noise signal formed by the primary sound source in the noise reduction region,thereby achieving the purpose of noise reduction.The main research contents of the project are as follows.Firstly,for the single noise situation in this experimental environment,the feedforward Wiener filter active noise control system structure is selected,and the current active noise control system is simplified,that is,the reference noise signal and the primary noise signal are recorded.The noise is provided through the computer's audio output.Secondly,in the secondary path identification,the NLMS algorithm is selected to identify the secondary path by comparing the convergence speed,convergence error and convergence time of the LMS algorithm,NLMS algorithm and RLS algorithm identification process.It was found in the experiment that the total number of recognition depends on the length of the reference white noise signal.In order to reduce the total number of identifications,the identification algorithm is improved.When the error is continuously smaller than a certain small value in the process of identification,it is considered that the identification process has reached convergence and the identification is terminated.Thirdly,in the active system,the core of the controller is the design of the filter.In view of the large difference in noise reduction between two error microphones in a two-channel system,another filter design is adopted to improve the robustness of the active control system.Think of the data of the two-channel system measured twice as afour-channel system.According to the solution process of the four-channel system,the weight of the filter is obtained,and the solution of the two-channel system is extracted.Finally,the feasibility of the two-channel active noise control system algorithm is verified by specific experiments.In the experiment,narrow-band white noise below 700 Hz,a piece of news audio with low frequency components and narrow-band white noise below 1500 Hz were used to design the weight coefficients of the filters in the controller,and the respective filter weight coefficients were recorded for different noises.The resulting theoretical noise reduction effect and actual noise reduction effect.Through experimental data,it can be found that the filter weight coefficient designed by narrow-band white noise below 700 Hz can generate more than 10 dB of noise reduction for narrow-band white noise,and generate noise reduction of more than 7 dB and more than 9 dB for news audio and music,respectively.At this time,the control effect of the system on noise is significantly better than that of news audio and narrow-band white noise below 1500 Hz to design the filter weight coefficient.