| Acoustic emission is a nondestructive testing technology widely used in the fields of transportation,aerospace,new energy materials and so on.With the wide application of acoustic emission detection technology,the processing and analysis of acoustic emission signals only through parameter analysis can not meet the current research needs.Using waveform analysis to process the acoustic emission signal and obtain the characteristic law of the signal has become a new technical means and research direction.Based on the acoustic emission signal of rock fracture,this paper processes and analyzes the collected signal by using the combination of S-transform and SVD,and uses the method of all phase difference time delay estimation,which provides strong theoretical support for the further study of rock fracture law and has a wide application prospect.The specific research work and achievements of the subject mainly include the following aspects:1.In this paper,through the indoor rock uniaxial fracturing experiment,sensors are arranged around the rock sample.The acoustic emission signals generated in the process of rock fracturing are collected by the sensors,and then the acoustic emission signals during rock fracture are obtained by the acoustic emission detection system.On this basis,it provides experimental data for subsequent signal processing and research.2.For the rock fracture signal obtained from the acoustic emission detection device,STSVD,a method combining S-transform and SVD,is used to denoise and reconstruct the signal.The singular value threshold determined by the singular value ratio method is compared with the singular value threshold selected based on K-means,and the method with good denoising and reconstruction effect is selected to achieve the purpose of denoising and reconstruction of rock fracture acoustic emission signal.The above two methods based on different singular value threshold are used to process the acoustic emission signals of rock fracture.By comparing their signal-to-noise ratio and mean square error,a better denoising effect is obtained.Through simulation analysis,it is concluded that selecting singular value threshold based on K-means is better for processing Rock signals,and provides strong support for the next acoustic emission source location.3.By means of generalized S-transform,the acoustic emission signals of rock fracture with different contents of brittle minerals are simulated and analyzed,and the variation law of acoustic emission signals under external loading of rock mass is studied.The analysis shows that although the content of brittle minerals in rocks is different,the main frequency is mostly between 0～200k Hz.With the increase of the content of brittle minerals in rocks,the concentrated frequency band decreases.4.Since the determination of the specific position of the acoustic emission source through the threshold acquisition time will lead to a great error,it is considered that the time can be obtained from the perspective of the waveform of the signal itself.This paper studies the all phase FFT spectrum analysis method,deduces its derivative formula and compares it with the traditional FFT,and obtains the "phase invariant" characteristic,that is,to solve the problems of spectrum leakage and phase spectrum disorder of the traditional FFT,so as to obtain a more accurate phase of its signal.Then the time delay between the current sensors can be calculated from the phase difference between different sensors.Compared with the system’s own localization,the acoustic emission source localization based on all phase FFT spectrum analysis has higher accuracy and lower mean square error. |
