As a real-time,dynamic,and continuous monitoring method,microseismic monitoring technology has been widely used in mining,petroleum,and geotechnical engineering.Coal mine microseismic signals contain rich information about coal and rock mass fractures,and are widely used in microseismic positioning and rockburst monitoring.During mine microseismic monitoring operations,a large number of signals are received,and only a small portion of these are microseismic signals.The common types of seismic sources in the underground can be generally divided into four categories: noise generated by the operation of mechanical equipment and electrical circuits,mining signals generated by equipment such as shearers and roadheaders,blasting signals and trough wave signals generated by blasting,and microseismic signals generated by coal and rock mass fractures and collapses.The characteristics of these sources can be obtained from three major categories: frequency domain,time domain,and time frequency domain.The main research results are as follows:A method of modifying the spectrum and removing invalid information through median filtering and peak fitting is proposed in the frequency domain.Only the main frequency,high and low cutoff frequencies,and other features are retained in the modified spectrum.It can accurately read the characteristics of the signal in the frequency domain in the modified spectrum,and perform automatic filtering based on the high and low cutoff frequencies of the signal.The effect of automatic filtering is better than that of fixed frequency filtering.A method for automatically picking up first breaks in the time domain is proposed: traverse each wave peak forward from the maximum value.When the ratio of noise amplitude to signal peak amplitude is greater than a certain coefficient,the wave peak is a first break.The value of this coefficient is inversely correlated with the time domain signal-to-noise ratio,and the relationship curve between this coefficient and the time domain signal-to-noise ratio is given.Based on the results of first break picking,characteristics in the time domain such as rise time,fall time,and signal envelope are obtained.In the time and frequency domain,it is proposed to use short-time Fourier transform to obtain the accuracy of the first break pickup of the signal,as well as the differences in the longitudinal and transverse wave distribution and frequency components.Based on the geological conditions and underground observation system of Zhaogu No.1 Mine,a physical similarity model is built using sand,gypsum,lime,etc.,and a similar observation system is built on the model.Simulate the actual mining process under the mine by excavating the model,and simulate the microseismic source and blasting source on the model.The signal simulated on the model and the signal actually mined under the mine have similar characteristics in various aspects.The differences between noise,mining signals,microseismic signals,and blasting signals are analyzed from three perspectives: frequency domain,time domain,and time frequency domain,and the methods for distinguishing different types of signals are summarized.It is found that at least one of the time-domain and frequency-domain signal-to-noise ratios of noise is very low.The dominant frequency of mining signals is concentrated on several fixed frequencies.The dominant frequency of blasting signals is higher,the duration is shorter,and the frequency component difference is large.The dominant frequency of microseismic signals is lower,the duration is longer,and the frequency component difference is small. |