Research On The Identification Method Of Microseismic Monitoring Waveform Category In Dahongshan Iron Mine

Underground mine microseismic monitoring signal contains information such as rock rupture,blasting,artificial activity,and noise.The recognition of rock rupture signal types is the premise and basis for rockburst prediction,goaf collapse prediction,and rock rupture location,and the accuracy of signal recognition has an important influence on the accuracy of judgment.This paper takes the microseismic signal of Dahongshan Iron Mine as the research object.On the basis of theoretical research and site investigation,multiple theoretical methods are first applied to the original signal such as Wavelet threshold transform,Empirical Mode Decomposition(EEMD),Hilbert-Huang Transform(HHT)and fractal theory and so on,and these methods have been used to extract the signal eigenvalues through theoretical analysis,theoretical programming calculation,engineering testing and other research methods and means.And then,after analyzing the signal eigenvalues in time domain,frequency domain,energy distribution and linear characteristics,a signal recognition model is established to recognize the two types of signals.The recognition of rock rupture and rock blasting signals is realized,which provides a basis for timely,effective and accurate recognition of rock rupture information in Dahongshan Iron Mine,and also provides a reference method for similar mines to identify the types of microseismic signals.The main contents of this paper are as follows:(1)Analyze the time-frequency characteristics,energy distribution characteristics,and linear characteristics of the signal to obtain the eigenvalues of the signal in terms of duration,amplitude,main frequency,signal energy distribution,fractal dimension,etc.law.(2)Based on multiple parameters such as duration,maximum amplitude,dominant frequency,energy distribution characteristics,and fractal dimension values of two types of signals,BP neural network prediction models for two types of event signals of rock failure and blasting were established.And the prediction ability of the model is verified.(3)In signal pattern recognition,the obtained characteristic parameter values are combined with MATLAB programming,and the BP neural network model is used to identify the 20 microseismic event signals collected on the site of Dahongshan.The results show that the model can effectively The above two types of signals are identified.