Microseism signal arrival time difference automatic picking method

Abstract

The invention discloses a microseism signal arrival time difference automatic picking method. The method comprises the following steps that microseism signals are collected by two sensors, the microseism signals are subjected to cross wavelet transform, results of cross wavelet transform are divided into an energy spectrum and a phase spectrum, the energy spectrum and the phase spectrum are self-coded, data is input into a neural network for calculation, and the result is the final microseism signal arrival time difference picking result. When the arrival time difference is calculated by the method, the arrival time does not need to be picked up independently for difference making then, two times of error introduction of independent arrival time picking can be avoided. Similar properties of homologous signals are adopted, the advantages of deep learning in extracting image feature information are adopted, and therefore signal similarity is matched, the homologous signals are accuratelyidentified, and the more accurate arrival time difference can be calculated. The method is high in reliability, high in accuracy, and is convenient and simple.

Description

technical field [0001] The invention specifically relates to a method for automatically picking up the time difference of microseismic signals. Background technique [0002] With the development of economy and technology, people's understanding and research on nature are getting deeper and deeper. [0003] Microseisms are small vibrations produced by rock fractures or fluid disturbances. In a broad sense, microseisms can be divided into two categories: microseisms in engineering production and microearthquakes in nature. The study of microseismic is helpful to people's understanding and research on earthquakes and artificial vibrations, so the study of microseismic is extra important. [0004] Microseismic records are characterized by high frequency and low signal-to-noise ratio, so the automatic identification of microseismic events and first-arrival time picking are of great significance for the automatic processing of massive microseismic data. For natural earthquake e...

AI Technical Summary

Problems solved by technology

Compared with the records of natural earthquakes, the magnitude of microseismic events is smaller, usually below magnitude 1, and the signal-to-noise ratio is lower, so it is more difficult to identify microseismic events

[0005] The currently commonly used microseismic signal-to-time difference recording method has poor recording accuracy and is no longer suitable for more and more in-depth and precise microseismic research.

Images