A Recognition Method for Microseismic and Blasting Events Based on Waveform Image

Abstract

The invention discloses a method for identifying microseismic and blasting events based on waveform images, which includes the following steps: Step 1: Establishing a waveform image database of microseismic and blasting events; Step 2: Acquiring image features of various events: using principal component analysis ( PCA) Extract the original waveform image features for M groups of microseismic events and N groups of blasting events, then reduce the data dimension and eliminate the correlation between different features, while quantitatively retaining the most useful feature information contained in the original waveform image features; Step 3 : Establishment of the classification model: use the machine learning algorithm LIBSVM to train the acquired waveform image features after dimensionality reduction, and establish a classification model for microseismic and blasting events; Step 4: Identify the event to be identified: input the waveform image feature of the event to be identified, According to the established classification model of microseismic and blasting events, the events to be identified are classified. The invention has the characteristics of wide application, accuracy and speed, strong objectivity and the like.

Description

technical field [0001] The invention belongs to the field of microseismic monitoring, in particular to a method for identifying microseismic and blasting events based on waveform images. Background technique [0002] Microseismic monitoring, as an effective means of ground pressure monitoring, has been widely used at home and abroad, and the identification of microseismic and blasting events is a key issue in the data processing process, so the research on the identification method of microseismic and blasting events is of great significance. However, for a normal working microseismic monitoring system, the number of microseismic and blasting events monitored every day is extremely large, and there are similarities between different events. The traditional manual identification method is easily affected by the operator's professional knowledge and experience. Not only is the number of identifications limited, but it may also cause inconsistent identification results, resulti...

AI Technical Summary

Problems solved by technology

There are a large number of commonly used source parameters, and more source parameters can be derived by fitting or converting different functions. It is extremely time-consuming and labor-intensive to select source parameters with better identification ability through statistical analysis

At the same time, the combination of different seismic source parameters with better identification ability also needs multiple tests and verifications, and is likely to be affected by experience and subjective judgments

For waveform parameters, usually only the front waveform is considered, and the positive effect of the full waveform on the identification of microseismic and blasting events is ignored

In addition, the importance of the selected waveform parameters is not included in the process of establishing the classification model, and the correlation between different parameters is not eliminated, which affects the classification effect

[0004] It can be seen that the existing identification methods for microseismic events and blasting events have relatively large limitations, and it is necessary to study an automatic identification method with strong applicability and high accuracy

Images