Research On Acoustic Emission Recognition Technology Based On Compressed Sensing

With the continuous development of information and science technology,people have become increasingly demanding for information security.Acoustic emission(Acoustic Emission,AE)recognition technology is widely used in basic building materials and a variety of equipment.A complete acoustic emission recognition system consists of pre-processing,feature extraction,pattern matching and determining.Only the acoustic emission signal can be accurately identified in order to further study the acoustic emission signal.At the same time,most of the acoustic emission signals contain noise in reality,so it is urgent to find a robust acoustic emission recognition technology.On the other hand,in recent years,the theory of compression sensing has been widely used,and its advantage is that the sampling rate of the signal is much lower than the Nyquist frequency,and the signal is sampled and compressed at the same time.This gives a huge advantage when acquiring large amounts of data.In this paper,the combination of compression sensing theory and acoustic emission recognition technology,this breakthrough enhances the performance of acoustic emission recognition system,and it is a significant research in practical application.In this paper mainly research compressed sensing theory and acoustic emission recognition technology,I have done the following work:1)The characteristics of acoustic emission and its application are studied in detail.The detection technology and processing method of acoustic emission signal are discussed,and the method of feature extraction of acoustic emission signal is studied.2)This paper studies the application of sparse representation theory in pattern recognition,analyzed the relationship between the theory and compression sensing,and proposed that a method of constructing sparse base based on GMM mean supervectors.Then,an acoustic emission identification algorithm based on compressed sensing is proposed,and the combination of the two theories is of great research value.3)In order to obtain the sparse solution of the signal,the application of the compression perception based on the matching tracking algorithm in the acoustic emission recognition technology is studied deeply.The convergence speed of the algorithm is fast,and the idea is simple,the recognition rate of the acoustic emission signal is improved effectively.4)After in-depth study of the compressed sensing algorithm under the Bayesian theory,The algorithm can obtain the sparse solution under the influence of observation noise.In view of the fact that the algorithm gets the solution closer to the minimum 0-norm solution,and first proposed acoustic emission recognition algorithm based on Bayesian compressed sensing.Finally,this paper also summarizes the research and achievements,and also prospects the future work.