Research On Intelligent Pickup-to-Time Method For Microseismic

Microseismic technology is an important tool used for safety and stability monitoring of disaster rock masses,which can effectively avoid the occurrence of coal mine power disasters.The core of microseismic technology is the location of seismic source,and the accurate and efficient vibration wave arrival time pickup directly affects the accuracy of microseismic location,the traditional manual arrival time pickup method is inefficient,while the automatic pickup method is affected by the parameters and cannot meet the arrival time pickup of massive microseismic event data.Therefore,finding a stable,fast and intelligent method to pick up vibration waves with high accuracy is a major problem in the field of microseismic monitoring.This paper is supported by the Zhangjiagang Industry-Academia-Research Advance Research Fund Project(No.ZKCXY2112),which aims to realize the intelligent pickup of vibration wave arrival time.-net vibration wave arrival time intelligent pickup algorithm,verify its feasibility by laboratory acoustic emission test,test the stability of the algorithm by arranging micro-earthquake monitoring system in the coal mine site,locate the arrival time results picked up by the algorithm,analyze the causes of the formation of large energy events in the working face,and give corresponding prevention and control suggestions,and achieve good results.The research content is as follows:(1)Replicate the traditional vibration wave arrival pickup algorithm and U-shaped convolutional neural network algorithm,analyze the characteristics of the algorithms,try to combine the two,improve the U-net code so that it can be used for vibration wave arrival pickup,and propose an intelligent pickup arrival method based on U-net.(2)Through laboratory acoustic emission test,acoustic emission data of coal rock and mortar specimen rupture were obtained and input into the U-net model for training,and the trained model,traditional AIC algorithm,and time-window energy eigenvalue algorithm were used for acoustic emission time-to-arrival pickup,and the arrival time accuracy was calculated,and it was concluded that the accuracy of U-net was slightly lower than that of AIC algorithm and higher than that of time-window energy eigenvalue algorithm;the stability was higher than that of AIC The conclusion that the stability is higher than that of the AIC algorithm and the time-window energy eigenvalue algorithm proves the feasibility of the U-net model for acoustic emission data arrival time pickup,and the reliability of the U-net model and the AIC algorithm is further verified by finding that the localization results of the U-net model and the AIC algorithm are closer to the actual rupture situation through source localization.(3)Taking the 13010 working face of Hengtai coal mine as the engineering support,arranging microseismic monitoring equipment,acquiring microseismic data,conducting network training,testing the accuracy of the algorithm,using the arrival time picked up by U-net for positioning,drawing a 3D working face positioning map,counting the energy distribution,and conducting disaster warning analysis,concluding that the engineering application accuracy of U-net algorithm is lower than the test accuracy,and the positioning part is in line with the actual conclusion of the situation...