Time-frequency peak filtering microseismic data random noise suppression method

Abstract

The invention relates to a time-frequency peak filtering micro-seismic data random noise suppression method, which comprises the following steps of: 1) reading one-dimensional micro-seismic data and decomposing the one-dimensional micro-seismic data by adopting a CEEMDAN algorithm to obtain modal components IMFs; 2) calculating a sample entropy (SE) of each IMFs; 3) dividing a noise dominant component and an effective signal dominant component according to the sample entropy; performing TFPF filtering on the noise dominant component, and reserving an effective signal dominant component; and 4) reconstructing the component filtered by the TFPF and the reserved component to obtain micro-seismic data after random noise suppression. The method has the advantages that in the denoising process, the TFPF window length is flexibly selected for IMFs needing to be filtered, random noise can be effectively suppressed, meanwhile, the effective signal amplitude can be well protected, effective micro-seismic signals can be accurately recognized, the positioning precision can be improved, the data characteristics of original micro-seismic events can be reserved, and reliable basic data can be provided for subsequent micro-seismic data research.

Description

technical field [0001] The invention belongs to the technical field of microseismic data processing, and in particular relates to a method for suppressing random noise of microseismic data based on CEEMDAN time-frequency peak filtering. Background technique [0002] Microseismic monitoring technology is a geophysical real-time detection technology that uses microseismic signals generated during rock mass rupture to study and evaluate rock mass stability in geotechnical engineering. By analyzing the microseismic signals generated in the process of rock mass damage and rupture, the source of microseismic events is located, and the internal stress distribution state of the surrounding rock in the stope, the law of roof activity, the spatial fracture form, and the release of energy are monitored. [0003] The effective signal energy in microseismic data is weak, the noise types are complex, and its signal-to-noise ratio is much lower than that of conventional seismic records. Ho...

AI Technical Summary

Problems solved by technology

[0007] Aiming at the problems existing in the suppression of random noise of microseismic data, the object of the present invention is to provide a method for suppressing random noise of microseismic data with time-frequency peak filtering. It is decomposed into a series of IMFs, and then according to the SE of each IMFs, the noise-containing signal is divided into a signal-dominant part and a noise-dominant part, and an adaptive window length adjustment scheme is constructed for different IMFs. The signal-dominant part adopts a shorter filtering window length to protect the effective signal Amplitude, the noise dominant part uses a longer filter window length to better eliminate noise, so that the TFPF algorithm adaptively realizes the balance between signal retention and random noise suppression, and provides an effective way to suppress random noise in microseismic data. Methods

Images