Research On Sound Source Recognition Method Base On Moving Microphone

Sound source recognition has many applications in many fields,such as noise control,fault diagnosis,voice input and so on.The commonly used sound source recognition methods are based on microphone array.Microphone array is used for sound source recognition.Due to the time difference of the sound source signals received by each element in the array,the sound source recognition is completed based on the time delay information.When a single mobile microphone collects the sound source signal in motion and segments the signal,there is also a time delay between each segment of the signal.Based on this,the correlation of recognition is established to complete the sound source recognition.In this paper,based on the single moving microphone sound source method,the influence of the number of segments on the sound source recognition is studied.When the distance between the array elements is larger than one wavelength of the sound source,the ghost phenomenon exists and interferes with the recognition of the sound source.At this time,the distance between the microphone array should be adjusted to make it less than one wavelength of the sound source to meet the recognition requirements,but the adjustment process of the array is more complicated.In the identification method of a single moving microphone,the time-domain signal needs to be segmenting.If the moving distance of the microphone is greater than one wavelength of the sound source within the segmenting time,the ghost phenomenon will also appear.At this time,the number of segments need to be adjusted to make the moving distance of the microphone less than one wavelength of the sound source.On this basis,it is found that when the phase difference between adjacent segments is less than 90 degrees,the sound source recognition effect is the best.Microphone moving distance of the results,the influence of noise source identification,in the quiet environment,microphone moving short distance can identify the sound source,when moving microphone in noise environment to identify the sound source,when the microphone moving distance is short,less amount of signal collected,the noise interference of sound source identification process,results there is a big error,noise source identification,With the increase of the moving distance of the microphone,the noise has less interference to the process of sound source identification,and the result of sound source identification is more accurate.Based on the recognition of plane sound source,the recognition of space sound source is studied.In the identification of spatial sound source,the motion track of moving microphone is plane motion,so the identification method of circular motion of microphone is put forward,and the simulation example shows that the circular motion of microphone can identify spatial sound source.Based on the plane motion of microphones,a method to identify spatial sound source with two moving microphones is proposed.The motion track of the two moving microphones is also plane motion,which can be used to identify spatial sound source.Based on two moving microphones,a spatial sound source identification method with three moving microphones is proposed.The motion track of the three moving microphones is spatial motion,and the spatial sound source can also be located.Study the space sound source identification method of three kinds of exercise antinoise performance,in the noise environment,when the acoustic source moving microphone distance is relatively close,three ways of sound source identification method can accurately identify the space position of sound source,the distance is not increase as the source depth,sound source identification method of planar motion is unable to accurately identify the sound source,The sound source recognition method with spatial motion is more sensitive to the depth of the sound source and the accuracy of the sound source recognition is more accurate.Based on the study of plane single point source and space single point source,the recognition of multiple point source is studied.An iterative subtraction multi-point sound source identification method is proposed.This method realizes the recognition of multiple point sound sources through continuous stripping and subtraction of multi-point sound source signals.The iterative subtraction identification method can be used to locate the multi-point sound sources with a long distance.However,the iterative subtraction identification method can not eliminate the mutual interference between the sound sources with a close distance,and can not identify multiple point sound sources.In this paper,Schmidt multi-point sound source recognition method is proposed.Multiple point sound sources are recognized by projection calculation of multi-point sound source on orthonormal basis.After the theoretical model is built,the theoretical model is tested and verified in a semi-anechoic chamber.The results show that the single mobile microphone sound source identification method can identify the sound sources of different frequencies and different depths with good results,which has a certain engineering value.