Acoustic Emission Signal Processing And Research Based On FRFT And K-means Algorithm

Rock fracturing refers to the formation of a connected fracture network inside the rock under the influence of external force,which collects the oil and gas energy dispersed in the rock to the local area,and forms the purpose of development within the scope.At present,rock fracturing technology is the main technical means for oil and gas energy development,and a lot of useful information contained in the rock fracturing process can be reflected by the acoustic emission signals generated during the fracturing process.Therefore,this paper mainly uses the acoustic emission signal generated by indoor rock fracturing as the carrier,and uses fractional Fourier transform,time-frequency analysis method,particle swarm algorithm and K-means algorithm to analyze the collected acoustic emission signal.Noise and feature analysis to study the corresponding relationship between the acoustic emission signal generated during rock fracture and rock fracture cracks.The specific research work and research results of this topic include the following four aspects:1.Design the rock fracturing experiment.Acoustic emission signals generated in the process of rock fracture are collected by the acoustic emission detection system and rock fracture loading system,and the signal waveform is displayed by the AEWIN visualization software;based on the collected acoustic emission signals,experimental data are provided for the following theoretical analysis.2.The collected acoustic emission signal is denoised.First,the optimal order of the rock fracture acoustic emission signal under different brittle mineral content is determined by the plane peak search method and the sparse metric method;then different types of discrete fractional Fourier transform(DFRFT,Discrete fractional Fourier transform)The acoustic emission signals of rocks with different brittle mineral contents under the order are denoised.By comparing the signal-to-noise ratio and root mean square error of the signals before and after denoising,the most suitable DFRFT algorithm for rock fracture acoustic emission signals with different brittle mineral contents is selected.The best denoising algorithm can provide high-quality acoustic emission signal samples for the following signal feature extraction.3.Divide the different stages of rock fracture process.According to the release level of the signal energy and the change of the accumulated energy of the acoustic emission signal,the rock failure process is divided into four stages,namely the elastic stage,the micro-crack initiation stage,the macro-crack continuous appearance stage,and the rock instability failure stage.The correlation analysis of duration is used to further analyze the acoustic emission signal characteristics of the four stages of the rock fracture process.4.Determine the type of fracture produced by the rock fracture process.Firstly,based on the analysis law of rock rupture acoustic emission signal,a rock rupture acoustic emission signal identification model with K-means algorithm as the core is established,and particle swarm algorithm is added to optimize the identification model of K-means algorithm;then the established acoustic emission signal identification model The main characteristics of the signal are statistically analyzed,and the fracture type corresponding to the acoustic emission signal is basically determined;finally,the four stages of the rock fracture process are effectively identified by combining the cumulative impact times and cumulative count values,and the rock fracture mode is analyzed.Process evolution,determine the rupture type of the acoustic emission signal generated during the rock rupture process,and further verify the reliability of the signal identification model.