Comprehensive Early Warning Method Of Gas Outburst Based On Microseismic Dynamic Response Law And Its Application

Coal and gas outburst（hereinafter referred to as gas outburst）is one of the most serious disasters in coal mines,which seriously threatens the safe production and energy security of coal mines in China.The occurrence conditions of coal seams in Guizhou are complex,the structure is changeable,and the gas outburst disasters are frequent,especially in geological abnormal areas such as structure and stress concentration areas.With the gradual increase of the mining depth of the mine,the hazard of gas outburst is becoming increasingly serious and complex.It is urgent to study the real-time monitoring technology of gas outburst and improve the accuracy of early warning,which meets the national requirements.However,it is difficult to retrieve the evolution process of gas outburst and dangerous area by existing technical means,and further in-depth research is needed in such aspects as real-time monitoring of outburst,location analysis of dangerous area,and effective early warning.Microseismic monitoring can realize the source location and reflect the regional fracture evolution process.It has the advantages of wide monitoring range and spatial positioning.It has a good development prospect in gas outburst monitoring and early warning,and is expected to solve the above problems.In view of the needs and difficulties,this paper adopts the research methods of theoretical analysis,numerical simulation,laboratory test and field test and verification to study the microseismic dynamic response and precursor quantitative characterization of the evolution process of gas outburst,the dynamic evaluation and classification of gas outburst risk areas,and the intelligent early warning method of gas outburst risk based on multi-source data fusion.The main research work and results are as follows:（1）The experimental system of coal body fracture,gas outburst and microseismic response is established.The experimental test and analysis reveal the microseismic dynamic response law of the process of coal and rock mass failure and instability,and a description method of gas outburst precursor characteristics based on microseismic indicators is proposed.Triaxial compression,confining pressure relief and gas outburst tests were carried out,and the microseismic monitoring data of the whole process of the test were obtained.Combined with the mechanical action process of the gas outburst,the microseismic dynamic response law and characteristics of the coal and rock mass failure and instability process were analyzed and excavated,and a quantitative description method of the gas outburst dynamic process based on the parameters of microseismic energy,amplitude and frequency was proposed.The research results show that the four stages of gas outburst（preparation,initiation,development and termination）have obvious characteristics of microseismic response,and their evolution law is significantly related to the energy,amplitude,frequency and waveform characteristics of microseismic events:（1）time-frequency characteristics:"energy reduction,frequency rise",quiet period is concentrated in the low frequency band,and the frequency distribution range of the outburst initiation and occurrence stage is wide and shifts to the high frequency,highlight the low noise feature after completion;（2）Phase characteristics:From the perspective of frequency,energy and other indicators,the large energy events in the early stage of the outburst gradually increased,and the large amplitude（energy）events increased at the time of the outburst,and the amplitude was generally much higher than the amplitude in the quiet period,while the event frequency increased sharply to the peak.（2）In view of the difficulties of complex and changeable microseismic signals in laboratory and field tests,difficult to identify and low positioning accuracy,a typical microseismic waveform library is built based on the measured data,and FSVD（Singular Value Decomposition of Frequency Domain）denoising method,HMM waveform recognition model and other algorithms are innovatively proposed,forming a microseismic data refinement processing technology that integrates"preprocessing-arrival time pickup-waveform recognition-precise positioning".Combining with the characteristics of"low energy,wide frequency band and low signal-to-noise ratio"weak signal,the FSVD microseismic data denoising method is proposed.The measured FSVD method ensures the improvement of time-frequency resolution and effectively eliminates false low-frequency interference components.After denoising,the signal trend is consistent with the original signal,and the signal-to-noise ratio is increased from less than 5db to more than 10db,and the original energy distribution is maintained;The extraction methods of amplitude,time and frequency domain characteristics of mine microseismic waveform are established,and the recognition system of mine microseismic waveform based on HMM model is constructed,with an effective recognition rate of 90%;An algorithm for picking up the initial arrival time of microseismic waveform is proposed,and the arrival time error after denoising by the above method is reduced by more than 70%.Based on this,a microseismic optimal positioning model combining simplex method and K-means clustering algorithm is constructed,and the relevant program is compiled using MATLAB to realize the rapid and high-precision positioning of regional microseismic monitoring（within the accuracy of 10m）.The above research laid a foundation for the quantitative characterization of coal and rock damage and the analysis of gas outburst precursor characteristics.（3）The microseismic response law of geological anomaly areas such as the fault layer during mining is revealed,the identification model of typical geological anomaly bodies based on clustering algorithm is established,and the detection and dynamic identification method of geological anomaly bodies based on microseismic is proposed.Taking a gas outburst mine in Guizhou Province as an engineering site,the microseismic response law and stress manifestation characteristics of mining face were studied,and the microseismic activity law during roadway excavation was obtained;This paper analyzes the microseismic activity in the geological anomaly areas such as faults and folds that have been exposed during mining,and puts forward the activation law and description method of geological anomaly areas based on microseismic indicators;An analysis method of microseismic activity characteristics in geological anomaly region based on DBSCAN clustering algorithm is proposed,and the automatic recognition and 3D visualization program of typical geological anomaly body is compiled using MATLAB;Taking a fault as an example,relevant field verification tests were carried out.The results show that the above method can obtain the spatial shape（position,dip angle,extension direction,etc.）information of the fault,and track the activation signs of the fault layer in the mining process in real time,which lays the foundation for the regional division and accurate early warning of coal and gas outburst disaster risk.（4）The microseismic response law and early warning field test in the process of gas outburst were carried out,the temporal and spatial evolution law of microseismic activity in the process of gas outburst disaster preparation was tested and analyzed,and the instability mode and microseismic precursor information of gas outburst disaster were revealed.Using the microseismic monitoring spatial positioning function,the real-time identification of potential dangerous areas is realized,the characteristics and distribution of stress and strain in the coal body are inverted,and the precursor characteristics of gas outburst are excavated.The research shows that there is the possibility of direct gas outburst caused by the large magnitude microseismic event（M>2.0）near the mining face,while the small magnitude event（M<1.0）is closely related to the abnormal gas emission;The distribution density of microseismic events in the region directly reflects the intensity of the internal activities of the media;In addition,conventional gas outburst sensitive indicators（S、K₁、Δh₂and q value）confirm the correlation between microseismic indicators（frequency,energy and spatial position）and the change of gas emission.This laid a foundation for the location and early warning of gas outburst risk areas.（5）The gas outburst early warning mode of"detection evaluation analysis early warning"is put forward,the evaluation method of gas outburst risk area based on the improved multi-index grey target decision model is established,the gas outburst intelligent early warning model integrating multiple parameters is constructed,and the corresponding gas outburst intelligent early warning platform is developed.A gas outburst classification model based on improved weighted grey target decision model is proposed.Compared with the measured results,the accuracy of classification evaluation is more than 80%;Based on the outburst risk degree obtained from the evaluation,the multi-source data fusion algorithm and analysis model are used to realize the fusion analysis and processing of gas,vibration and other multi-field data,and an intelligent gas outburst early warning model based on artificial intelligence algorithm is established.The accuracy of the model is 91%by using the confusion matrix analysis;A comprehensive early warning method for gas outburst based on multi-source monitoring data is proposed,and an intelligent recognition and early warning platform for gas outburst hazard area based on B/S architecture is developed and implemented.Through more than half a year of on-site testing and verification,the system software has effectively warned 23 times without missing.This study provides a new method for the real-time monitoring and early warning research of gas outburst disaster,and has important theoretical significance and practical value for the in-depth study of the disaster mechanism of gas outburst disaster,improving the accuracy of early warning,and ensuring the safety of mine production.