Efficient intelligent sensing and collecting method for micro-seismic signal detection

Abstract

The invention discloses an efficient intelligent collecting method for micro-seismic signals. The method comprises compressed sensing sampling of micro-seismic data, extraction of effective micro-seismic events in the compressed data and reconstruction of compressed data of the micro-seismic events. Specifically, all distributed acquisition nodes firstly perform compressed sensing on microseism data so as to obtain a measurement value of a microseism signal. Redundant data of the microseismic signals in the time domain are removed through compressed sensing, and subsequent unnecessary calculation expenditure is avoided. And then, the measurement value of the microseism signal is input into a pre-trained deep neural network, and then the identification of the effective microseism event is finished. And after the valid event is identified, the acquisition node transmits the compressed sampling data containing the microseism event to a data center. And finally, the data center performs approximate value optimization through singular value decomposition and a clustering method so as to continuously update the sparse basis of the seismic data, and uses an l1 norm spectrum projection gradient (SPGL1) algorithm to reconstruct the original micro-seismic event data. In conclusion, the novel data acquisition method can improve the data acquisition efficiency of the whole system from twoaspects of data recording and data transmission.

Description

technical field [0001] The invention relates to the field of micro-seismic signal detection method design, in particular to an efficient intelligent sensing acquisition method for micro-seismic signal detection. Background technique [0002] With the rapid development of artificial intelligence algorithms, the academic community is paying more and more attention to its application in the field of resource exploration. For example, in the "Global Engineering Frontiers" released by the Chinese Academy of Engineering in 2018, the intelligent and efficient exploration and exploitation of oil and gas resources are also listed as the frontiers of engineering research, and it is pointed out that intelligent collection, efficient transmission and intelligent analysis of data are future development. the trend of. Therefore, combining the relevant theories of artificial intelligence, it is of great significance to conduct research on acquisition methods that improve the use efficienc...

AI Technical Summary

Problems solved by technology

However, due to the diversity of seismic data features, most of the methods for applying compressive sensing to efficiently acquire seismic data in acquisition nodes use fixed transformation domains as the sparse domains of seismic data (such as Curvelet, Seislet, etc.) The intersection conditions are used as constraints for dictionary learning (such as principal component analysis, K-SVD, etc.), so it is difficult to achieve high-quality representations for different types of seismic data

Moreover, these methods do not fully consider the high-order combination of sparse domain features when reconstructing seismic data, and it is difficult to restore details in seismic data with high precision under the premise of using fewer feature components. Therefore, for high compression ratio seismic data The ability to refactor is not strong

Images