Research On Suppression Method Of Impulse Noise Cluster For Acousticseismic Coupling Data

The low-altitude target detection and identification technology based on acoustic-seismic coupling wave is an important means to realize the low-altitude target detection and identification at present.Because of its long-term,high concealment and strong anti-interference ability,it is suitable for a variety of complex environments.Based on the principle of acoustic-seismic coupling,the signal of acousticseismic coupling data is weak,but its noise interference is strong,mainly including impulse noise,random noise and power frequency noise.There are many literature studies on the suppression of random noise and power frequency noise,but there are few studies on the suppression of impulse noise of acoustic-seismic coupling data.Moreover,Due to the variety of impulse noise sources,impulse noise is ubiquitous and often appears in the state of aggregation,which has the characteristics of long duration,high amplitude and strong energy.The data quality of conventional noise suppression methods after noise suppression is poor,which is difficult to be used for low-altitude target detection and recognition.Therefore,in order to meet the demand for high-quality acoustic-seismic coupling data for low-altitude target detection and identification,this paper studies impulse noise suppression.With the support of the elastic wave detection scientific research project,this paper has carried out the research on suppression method of impulse noise cluster for acoustic-seismic coupling data.On the basis of analyzing the characteristics of the impulse noise of acoustic-seismic coupling data,the suppression method of impulse noise cluster for acoustic-seismic coupling data based on AT-STA/LTA(Adaptive Threshold Short-Term Average/Long-Term Average,AT-STA/LTA)and the suppression method of impulse noise cluster for acoustic seismic coupling data based on adaptive sensing of weak signal are proposed.Combined with the noise suppression results of simulation and measured data,the effectiveness and superiority of the above two noise suppression methods are verified.The main contents are as follows:Firstly,the impulse noise characteristics of the measured acoustic-seismic coupling data are analyzed.Based on the analysis of the duration of the impulse noise and the composition of the noise,it is obtained that the impulse noise of acousticseismic coupling data has the characteristics of "cluster".The specific analysis of the impulse noise cluster shows that it has the characteristics of short-term amplitude mutation,short-term frequency mutation and irrelevance to the effective signal,which provides a theoretical basis for the design of subsequent noise suppression algorithms.Based on the short-term amplitude and frequency mutation characteristics of noise,the paper proposes to use the STA/LTA(Short-Term Average/Long-Term Average,STA/LTA)method to pick up impulse noise,which is difficult to take into account both the wrong pick-up and noise reduction of noise.Therefore,the AT-STA/LTA method is proposed to achieve the accurate picking of impulse noise,and the adaptive median filtering method is combined to filter the picked-up noise.After the method is used to denoise the simulation and measured data,the signal-to-noise ratio of the simulation data is improved by about 9.24 d B,and the effective signal of the measured data is clearly distinguishable.After analysis,the noise suppression method based on AT-STA/LTA has the problem of signal distortion when dealing with low signal-to-noise ratio data with signal-to-noise aliasing.Therefore,based on the difference of signal-to-noise correlation,the paper constructs a noise suppression model based on signal adaptive cancellation,and proposes a noise suppression method based on weak signal adaptive sensing to achieve high data fidelity.Moreover,in order to improve the noise suppression performance of the method,a correlation delay estimation method based on AT-STA/LTA filtering is proposed.The application results of simulation and measured data show that the method has a significant denoising effect and can recover the weak signal submerged by the impulse noise cluster.