CN-121996922-A - Data denoising method, device, electronic equipment, storage medium and program product

CN121996922ACN 121996922 ACN121996922 ACN 121996922ACN-121996922-A

Abstract

The embodiment of the application relates to the technical field of data processing, and discloses a data denoising method, a device, electronic equipment, a storage medium and a program product, wherein the method comprises the steps of acquiring DAS data to be processed in a mining area environment; performing curvelet transformation on DAS data to be processed to obtain a curvelet coefficient set, wherein the curvelet coefficient set corresponds to a plurality of scales, the plurality of scales comprise at least one key scale, the at least one key scale is determined based on a frequency band of an effective signal in a mining area environment, the density of a plurality of directions corresponding to each key scale in a key direction range is higher than that in a non-key direction range, the key direction is determined based on geological features of the mining area environment, denoising is performed on the curvelet coefficient set to obtain a denoised curvelet coefficient set, and first DAS data is generated based on the denoised curvelet coefficient set. This way the quality of DAS data is improved.

Inventors

WANG BIN
LIU BINGXUAN
LI GUOQIANG

Assignees

中国神华能源股份有限公司哈尔乌素露天煤矿

Dates

Publication Date: 20260508
Application Date: 20251216

Claims (10)

1. A method for denoising data, comprising: acquiring to-be-processed distributed acoustic wave sensing DAS data in a mining area environment; Performing curvelet transformation on DAS data to be processed to obtain a curvelet coefficient set, wherein the curvelet coefficient set corresponds to a plurality of scales, the scales comprise at least one key scale, the at least one key scale is determined based on a frequency band of an effective signal in the mining area environment, the density of a plurality of directions corresponding to each key scale in a key direction range is higher than the density in a non-key direction range, and the key direction is determined based on geological features of the mining area environment; denoising the curvelet coefficient set to obtain a denoised curvelet coefficient set; and generating first DAS data based on the denoising curvelet coefficient set.
2. The method for denoising data according to claim 1, wherein the acquiring the to-be-processed distributed acoustic sensing DAS data in the mining area environment comprises: Acquiring original DAS data in the mining area environment, wherein the original DAS data comprises a plurality of time sequences; constructing a common mode noise vector based on a preset quantile of the amplitudes of the plurality of time series at each time point; determining cross-correlation coefficients between each time sequence and the common mode noise vector; And removing the product of the cross-correlation coefficient corresponding to the time sequence and the common mode noise vector from each time sequence of the original DAS data to obtain the DAS data to be processed.
3. The method for denoising data according to claim 1, wherein denoising the set of curvelet coefficients to obtain a denoised set of curvelet coefficients comprises: Denoising the curved wave coefficient set according to the following hard threshold function: Wherein, the Is a curvelet coefficient in the curvelet coefficient set, For the sum of the denoising curvelet coefficient set Corresponding curvelet coefficients; Is that 、 One of or is And (3) with Is a product of (2); is a geological weight determined based on geological features of the mining area environment, Is that Is a function of (a) and (b), Is the position; is a statistical weight determined based on the local mathematical statistical properties of the curvelet coefficients, Is that Is a function of (a) and (b), In order to be a scale of the dimensions, Is the direction; Is based on A base threshold for statistical distribution determination of wavelet coefficients within a subband.
4. The method of data denoising according to claim 1, wherein the generating first DAS data based on the denoised set of curvelet coefficients comprises: Acquiring the credibility weight of the sub-band corresponding to each scale-direction in the denoising curvelet coefficient set, wherein the credibility weight is determined based on the average energy of the sub-band corresponding to each scale-direction; weighting and adjusting the curvelet coefficient in the denoising curvelet coefficient set based on the credibility weight; and generating the first DAS data based on the denoised curved wave coefficient set after the weighting adjustment.
5. The method of data denoising according to claim 4, wherein generating the first DAS data based on the weighted adjusted set of denoised curved wave coefficients comprises: constructing constraint type: , wherein, For the preset reconstructed signal(s), Is that A set of curvelet coefficients after curvelet transformation, For the weighted and adjusted denoising curved wave coefficient set, The sign is calculated for the L2 norm, In order for the parameters to be regularized, To pair(s) Performing total variation treatment; aiming at minimizing the constraint, for the Optimizing, and completing the optimization As the first DAS data.
6. The data denoising method according to claim 1, further comprising: subtracting the first DAS data from the DAS data to be processed to obtain first residual data; Performing curvelet transformation and low-threshold denoising on the first residual data, and generating a first effective signal based on the obtained result; superposing the first effective signal and the first DAS data to obtain second DAS data; Subtracting the second DAS data from the DAS data to be processed to obtain second residual data; Performing similarity comparison on the second residual data and each effective signal in the effective signal mode library corresponding to the mining area environment, and taking the part, which is larger than a similarity threshold, of the second residual data and each effective signal as a second effective signal; Superposing the second effective signal and the second DAS data to obtain third DAS data; subtracting the third DAS data from the DAS data to be processed to obtain third residual data; Removing components lower than physical background noise in the mining area environment in the third residual data to obtain a third effective signal; And superposing the third effective signal and the third DAS data to obtain target DAS data.
7. A data denoising apparatus, comprising: The acquisition module is used for acquiring the to-be-processed distributed acoustic wave sensing DAS data in the mining area environment; The processing module is used for carrying out curvelet transformation on DAS data to be processed to obtain a curvelet coefficient set, the curvelet coefficient set corresponds to a plurality of scales, the plurality of scales comprise at least one key scale, the at least one key scale is determined based on a frequency band of an effective signal in the mining area environment, the density of a plurality of directions corresponding to each key scale in a key direction range is higher than the density in a non-key direction range, and the key direction is determined based on geological features of the mining area environment; the denoising module is used for denoising the curvelet coefficient set to obtain a denoised curvelet coefficient set; and the generation module is used for generating first DAS data based on the denoising curvelet coefficient set.
8. An electronic device, comprising: At least one processor, and A memory communicatively coupled to the at least one processor, wherein, The memory stores instructions executable by the at least one processor to enable the at least one processor to perform the data denoising method of any one of claims 1 to 6.
9. A computer readable storage medium storing a computer program, wherein the computer program when executed by a processor implements the data denoising method of any one of claims 1 to 6.
10. A computer program product comprising a computer program, characterized in that the computer program, when executed by a processor, implements the data denoising method of any one of claims 1 to 6.

Description

Data denoising method, device, electronic equipment, storage medium and program product Technical Field The embodiment of the application relates to the technical field of data processing, in particular to a data denoising method, a device, electronic equipment, a storage medium and a program product. Background The distributed optical fiber sensing technology obtains DAS (Distributed Acoustic Sensing, distributed acoustic wave sensing) data by measuring scattered light along an optical fiber by using the optical fiber as a sensing medium, and high-precision and high-resolution monitoring of a specific area is realized by the DAS data. In the seismic exploration process of the mining area, the topography is complex and changeable, the traditional seismic instrument layout cost is high, and the distributed optical fiber sensing technology utilizes the optical cable to realize large-scale DAS data acquisition, so that the problems of topography obstacle and high cost in the mining area layout survey line are effectively solved. However, the DAS data acquisition process is greatly affected by noise, resulting in lower data quality and affecting the accuracy of subsequent signal processing and interpretation. How to denoise DAS data to improve data quality is a challenge. Disclosure of Invention The application aims to at least provide a data denoising method, a device, electronic equipment, a storage medium and a program product, which can at least solve the problem of denoising DAS data and at least achieve the effect of improving the quality of the DAS data. In order to solve the technical problems, at least one embodiment of the application provides a data denoising method, which comprises the steps of obtaining Distributed Acoustic Sensing (DAS) data to be processed in a mining area environment, performing curvelet transformation on the DAS data to be processed to obtain a curvelet coefficient set, wherein the curvelet coefficient set corresponds to a plurality of scales, the scales comprise at least one key scale, the key scale is determined based on a frequency band of an effective signal in the mining area environment, the density of a plurality of directions corresponding to each key scale in a key direction range is higher than the density in a non-key direction range, the key direction is determined based on geological features of the mining area environment, denoising the curvelet coefficient set to obtain a denoised curvelet coefficient set, and generating first DAS data based on the denoised curvelet coefficient set. The core parameters (scale and direction) of the curvelet transformation are optimized based on the effective signal frequency band and the geological features of the mining area environment, so that the denoising tool is not universal any more, but is customized for the mining area, and the denoising accuracy is fundamentally improved. By identifying and encrypting the directional subbands in the critical directional range of the critical dimension, the resolution and fidelity of the effective signals propagating along the main geological structure of the mining area are greatly improved, and the effective signals with geological significance can be better reserved. The non-uniform direction division means that the calculated amount can be reduced in the direction of insensitive signals, so that the overall efficiency of the algorithm is improved on the premise of not sacrificing key information, and the method is more suitable for processing large-scale data generated by DAS. In some optional embodiments, the acquiring the Distributed Acoustic Sensing (DAS) data to be processed in the mining area environment comprises acquiring original DAS data in the mining area environment, wherein the original DAS data comprises a plurality of time sequences, constructing a common mode noise vector based on preset quantiles of amplitudes of the time sequences at each time point, determining cross-correlation coefficients between each time sequence and the common mode noise vector, and removing products of the cross-correlation coefficients corresponding to the time sequences and the common mode noise vector from each time sequence of the original DAS data to obtain the DAS data to be processed. Common mode noise caused by instruments and the like exists in DAS data, and the common mode noise is directly preprocessed aiming at inherent in the DAS system and affecting in-phase noise of all channels, so that follow-up denoising can be facilitated, and the quality of first DAS data after denoising is improved. Common mode noise is easy to be confused with effective signals in a sparse domain, the common mode noise is removed preferentially, the interference to subsequent curvelet transformation and threshold processing can be prevented, and the denoising effect and reliability of the sparse domain are improved remarkably. The common mode noise is estimated by using a preset fractional number instead