The Aplication Of Wavelet Analysis In Mine Acoustic Emission Data And Early Warning

Throughout the research and development status of the monitoring technology at homeand abroad, the application of basic research as the core of the dynamics caused by thedisaster site monitoring has made some progress, but the method is confined to thetraditional stress, displacement, energy monitoring and analysis, only limited to thetraditional a priori means of statistical methods and regression analysis methods formonitoring signal processing and intelligent recognition, similar to acoustic emissionnon-steady-state continuous signal, these methods show some limitations. reduce thereliability and practicality of the analysis results, so effective monitoring and prediction ofdynamic stability problems caused by mining is a very important research topic.In this paper, by using the parametric method and Wavelet-PM model we denoised andsmoothed the mine Acoustic Emission data which was processed by parametric method inorder to extract the main features. The main work of this paper as follows:(A) Participate in the collection of a coal mine B3gallery3face acoustic emissions data.With the help of sensors will be received by the sound waves into electrical signalsby using the computer system, the sound of the original data transmission.(B)Used the parameter analytic method to analyze the B3tunnel three working surfacesound firing data, formed may reflect the signal initial characteristic the curve, includingringing counting-time, energy-time, total energy-time, total ringing counting-timecurve, and used statistical methods and so on cluster analysis, regression analysis has madethe analysis to the data, gave the conclusion.The conclusion may the direct-viewingreflection make noise the firing data characteristic, but its reflection characteristic uncertainand actual match case.(C)Classical wavelet analysis method to analyze the acoustic emission data.(1) I write acomputer program, draw a scatter plot of mine acoustic emission data, it can reflect the important and necessary features of the sound wave.(2) classical wavelet analysis method(using haar wavelet basis, select a reasonable threshold) noise on the scatter plot smooth.Results can be seen, the classical wavelet method can remove some of the details of theunreasonable signal, highlight the main features of the data.(D)Created Wavelet-PM nonlinear partial differential equation model for acousticemission data for feature extraction. As can be seen, the model is able to effectively removenoises such as local details, highlight the main characteristics of the signal, which canprovide important support for risk early warning.(E)According to the above research, makes the forecast to the same day B3tunnel powerjitter situation.