Automatic identification method of rock failure events in tunnel micro-seismic monitoring

Abstract

The invention provides an automatic identification method of rock failure events in tunnel micro-seismic monitoring, and relates to the technical field of tunnel micro-seismic monitoring. The method includes the steps that first, a large number of tunnel micro-seismic monitoring waveforms with known types are used for establishing a waveform sample database, then effective information of the waveforms in the waveform sample database is analyzed statistically, and the length of a waveform sample for deep learning is determined; an identification model of the tunnel micro-seismic monitoring waveforms based on a deep convolutional neural network is established; the large number of to-be-identified micro-seismic waveforms are input into the waveform identification model, and waveform type identification results of the to-be-identified micro-seismic waveforms are output; and finally, the types of the micro-seismic events are determined according to the waveform type identification results.According to the automatic identification method of the rock failure events in tunnel micro-seismic monitoring, an original waveform of micro-seismic monitoring is directly identified, waveform feature extraction is not needed, the influence of improper eigenvalue selection on the signal identification accuracy is avoided, identification from the types of the micro-seismic monitoring waveforms tothe types of the micro-seismic events is realized, and the identification results can be directly used for rock burst micro-seismic early warning.

Description

technical field [0001] The invention relates to the technical field of tunnel microseismic monitoring, in particular to an automatic identification method for rock rupture events of tunnel microseismic monitoring. Background technique [0002] As mines, tunnels (roads) and other projects gradually extend to deep areas, dynamic disasters such as rockbursts occur frequently, and microseismic monitoring technology is gradually applied to tunnel rockburst monitoring. Different from mine microseismic monitoring, tunnel microseismic monitoring contains a large number of noise signals, including electrical noise, mechanical noise, etc., which are similar and intertwined with rock fracture signals. Therefore, tunnel microseismic monitoring needs to quickly and accurately identify rock fracture events. The premise of timely and accurate early warning of rockburst risks. [0003] At present, there are mainly methods for automatic identification of microseismic events: waveform spectr...

AI Technical Summary

Problems solved by technology

Specifically, a rock fracture event will trigger multiple microseismic sensors, and only when multiple sensors are triggered, can it be considered as an effective fracture event. Even for a small energy rock fracture event, it may contain a certain amount of noise waveforms, which determines The type of microseismic waveform and the type of microseismic event are completely different concepts, and the two cannot be equated, which affects the accuracy of rock fracture event identification to a certain extent.

[0005] It can be seen that the existing methods for identifying rock fracture events in tunnel microseismic monitoring still have relatively large limitations, and it is necessary to get rid of the limitations of eigenvalue identification and manual identification, give the relationship between microseismic waveform types and microseismic event types, and establish a method for quickly identifying rock fracture events

Images