A method for noise reduction of time-shift resistivity monitoring data in mines

By combining VMD and analytical trend correction, the problem of noise interference in mine time-shift resistivity monitoring was solved, achieving high-precision and rapid noise suppression and improving the monitoring effect of water hazard risks.

CN117972324BActive Publication Date: 2026-11-10XIAN RES INST OF CHINA COAL TECH & ENG GRP CORP
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202410070457.3
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-01-17
Publication Date
2026-11-10
Estimated Expiration
2044-01-17

AI Technical Summary

Technical Problem

Noise interference is severe in mine time-shift resistivity monitoring data, affecting the effectiveness of water hazard early warning. Existing methods are insufficient to effectively suppress noise interference during the mining process.

Method used

A combined approach of variational mode decomposition (VMD) and analytical trend correction is used to remove high-frequency components from the measurement data and to correct the voltage analytical trend based on the operating environment, thereby reducing noise interference.

Benefits of technology

It improved the accuracy of noise removal, shortened the processing time, and enhanced the monitoring effect of time-shifted resistivity water hazard risks.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117972324B_ABST
    Figure CN117972324B_ABST
Patent Text Reader

Abstract

A noise reduction method for mine time-shift resistivity monitoring data is proposed, which adopts a combination of VMD and analytical trend correction. VMD removes high-frequency components from the measurement data, and then the voltage amplitude signal is corrected according to the voltage analytical trend of the working environment. The noise removal accuracy is high and the processing time is short, so as to solve the noise suppression problem in the mining process and improve the monitoring effect of time-shift resistivity water hazard.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of geophysical exploration technology, specifically relating to a method for noise reduction of mine time-lapse resistivity monitoring data. Background Technology

[0002] Time-shift resistivity monitoring (TL-ERT) is an important means of making geological information hidden in mine water hazards transparent. However, during the monitoring process, it is easily affected by mining disturbances and strong electrical interference, resulting in a large amount of random noise in the collected data, which leads to poor application effect of this technology in mine water hazard early warning.

[0003] One of the primary challenges in mine time-lapse resistivity monitoring is reducing noise interference. Conventional mine resistivity methods involve measurements taken when mining activities have ceased, resulting in errors primarily caused by poor contact electrode coupling and acquisition system errors. These errors can typically be addressed using outlier correction and smoothing filtering. However, unlike conventional resistivity detection methods, mine time-lapse resistivity monitoring is conducted under mining disturbances. The electrical signals generated by motors, transportation systems, and vibrations within the mine severely interfere with effective signal extraction, making it impossible to utilize trend smoothing-based signal processing methods. Therefore, there is an urgent need to address noise suppression during mining operations to improve the effectiveness of time-lapse resistivity monitoring for potential water hazard risks. Summary of the Invention

[0004] To address the aforementioned problems, the purpose of this invention is to provide a noise reduction method for mine time-shift resistivity monitoring data. This method employs a combination of VMD and analytical trend correction. VMD removes high-frequency components from the measurement data, and then the voltage amplitude signal is corrected based on the voltage analytical trend of the working environment. This method achieves high noise removal accuracy and short processing time, thereby solving the noise suppression problem in the mining process and improving the monitoring effect of time-shift resistivity water hazard risks.

[0005] To achieve the above objectives, the technical solution adopted by the present invention includes:

[0006] A method for noise reduction of mine time-lapse resistivity monitoring data includes the following steps:

[0007] S0, acquire m full waveform signals data(t) from m measurement points;

[0008] S1, for the i-th full waveform signal data i (t) Perform VMD decomposition to obtain imf of k modal components. k The modal component set composed of (t)

[0009] Γ i (t)={imf1(t),imf2(t),…,imf k (t)};

[0010] Where t represents the sampling time, the measurement point is the observation point, and the i is initially 1;

[0011] S2, the set of modal components Γ obtained from S1 i (t) Perform FFT to obtain the frequency domain signal set.

[0012] Γ i (f)={imf1(f),imf2(f),…,imf k (f)};

[0013] Among them, imf k (f) represents imf k (t) represents the frequency domain signal, where f is the sampling frequency;

[0014] S3, setting the transmission frequency as f' in the resistivity monitoring project, and processing the frequency domain signal set Γ obtained in S2. i (f) Filtering is performed at the transmission frequency f' to obtain the filtered dataset Γ. i_filter (f'), the filtered dataset Γ i_filter (f') is the set of k component signals;

[0015] S4, Obtain the filtered dataset Γ i_filter The maximum amplitude of each component signal in (f') is used to obtain the component signal amplitude set.

[0016] Γ i_filter_max (f')={imf 1_filter_max (f'),imf 2_filter_max (f'),…,imf k_filter_max (f')}, compared to

[0017] Comparison i_filter_max The size of each element in (f') is used to obtain the component corresponding to the maximum value, which is the j-th modal component imf. j (t);

[0018] Among them, imf k_filter_max (f') represents the filtered dataset Γ i_filter The maximum amplitude value corresponding to the kth component signal in (f');

[0019] S5, with the j-th modal component imf j (t) is the boundary, in the modal component set Γ i Remove the components corresponding to high-frequency components from (t) {imf1(t),imf2(t),...,imf j-1 (t)}, thus obtaining the non-high frequency modal component set Γ i_delete (t)={imfj (t),imf j+1 (t),…,imf k (t)};

[0020] S6, the non-high frequency modal component set Γ obtained in S5 i_delete Adding the elements in (t) yields the reconstructed mine time-shift resistivity monitoring signal Γ. i_re (t);

[0021] S7, Γ obtained from S6 i_re (t) is subjected to FFT to obtain the corresponding frequency domain signal Γ. i_re (f'), and extract Γ i_re The amplitude of the signal at frequency f' is denoted as Γ. i_re_amp ;

[0022] S8, obtain the analytical voltage U at the i-th measurement point. i ;

[0023] S9, let i = i + 1, repeat S1-S8 on the data of the remaining m-1 measurement points to obtain the set Γ of the amplitudes of the m measurement points at frequency f'. all_re_amp ={Γ 1_re_amp ,Γ 2_re_amp ,…,Γ m_re_amp}, and the set U of analytical voltages at m measurement points. all ={U1,U2,…,U m};

[0024] S10, the corrected voltage signal U of the i-th measuring point is obtained through equation (1). out_i ;

[0025]

[0026] Among them, Γ re_i U represents the amplitude percentage of the i-th measurement point at frequency f', as shown in equation (2); the initial value of i is 1; U re_i The percentage of the analytical voltage at the i-th measurement point at frequency f' is represented by equation (3);

[0027]

[0028]

[0029] S11, let i = i + 1, repeat S10 for the data of the remaining m-1 measuring points to obtain the noise-reduced mine time-shift resistivity monitoring signal U. out_all U outall ={U out1 U out2 , ..., U outm}

[0030] Preferably, in S3, the frequency domain signal set Γ is filtered. i (f) Perform filtering, with the filter boundary being [f'-αf', f'+αf']; where α is the filter bandwidth factor.

[0031] Preferably, α is 0.5.

[0032] Preferably, the analytical voltage in S8 Where, ρ i I represents the formation resistivity at the i-th measuring point. i r represents the excitation current at the i-th measuring point during measurement. i This represents the straight-line distance between the observation point and the excitation point in the original measurement data. The excitation point represents the point where the signal is transmitted.

[0033] Compared with the prior art, the advantages of the present invention are:

[0034] This application discloses a noise reduction method for mine time-shift resistivity monitoring data, which adopts a combined method of VMD and analytical trend correction. The VMD removes high-frequency components from the measurement data, and then the voltage amplitude signal is corrected according to the voltage analytical trend of the working environment. The noise removal accuracy is high and the processing time is short. Attached Figure Description

[0035] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the following detailed description to explain the invention, but do not constitute a limitation thereof. In the drawings:

[0036] Figure 1 This is the first full-waveform signal data1(t) collected by time-shift resistivity monitoring in a noisy mine;

[0037] Figure 2 It is the signal obtained after performing VMD mode decomposition on data1(t);

[0038] Figure 3 The effect of using the EMD method to decompose the acquired signal;

[0039] Figure 4 The effect of using the VMD method to decompose the acquired signal;

[0040] Figure 5 It is the Γ1(f') signal obtained after performing an FFT transform on Γ1(t);

[0041] Figure 6 Γ is obtained by filtering the Γ1(f') signal. 1_filter (f') signal;

[0042] Figure 7 The reconstructed mine time-shift resistivity monitoring signal Γ is obtained. 1_re (t) signal;

[0043] Figure 8 The amplitude Γ of 10 data points at a frequency of 15Hz all_re_amp Signal;

[0044] Figure 9 It is the analytical solution voltage U at the corresponding positions of 10 data points. all;

[0045] Figure 10 It is the noise-reduced mine time-shift resistivity monitoring signal U out_all .

[0046] Figure 11 A comparison chart of signal errors after voltage trend correction. Detailed Implementation

[0047] The invention is not limited to the specific embodiments described below. All equivalent modifications made based on the technical solutions of this application fall within the protection scope of this invention.

[0048] Unless otherwise specified, all components and devices in this invention are based on components and devices known in the prior art.

[0049] The VMD disclosed in this application refers to variational mode decomposition, as detailed in the literature "Dragomiretskiy, K.; Zosso, D. Variational mode decomposition. IEEE Trans. Signal Process. 2014, 62, 531-544".

[0050] The EMD disclosed in this application refers to Empirical Mode Decomposition, as detailed in the literature "Huang, NE, Shen, Z., Long, SR, Wu, MC, Shih, HH, Zheng, Q., Yen, NC, Tung, CC and Liu, HH. "The Empirical Mode Decomposition and the Hilbert Spectrum for Nonlinear and Non-stationary Time Series Analysis.", Royal Society Proceedings on Math, Physical, and Engineering Sciences, Vol. 454, No. 1971, pp. 903-995.

[0051] Example

[0052] S0, acquire m full waveform signals data(t) from m measurement points. In this embodiment, there are 10 measurement points and 10 full waveform signals data(t).

[0053] S1, for the i-th full waveform signal data i (t) Perform VMD decomposition to obtain imf of k modal components. k The modal component set Γ (t) constitutes i (t)={imf1(t),imf2(t),…,imf k (t)};where t represents the sampling time, the measurement point is the observation point representing the point where the signal is collected; the initial value of i is 1;

[0054] Variational mode decomposition (VMD) was performed on the first acquired full-waveform signal data1(t) to obtain six IMF(t) components containing different modes, denoted as Γ1(t)={imf1(t),imf2(t),imf3(t),imf4(t),imf5(t),imf6(t)}. The decomposed data is as follows: Figure 2 As shown; by decomposing the acquired signal using the EMD and VMD methods respectively, the following results can be obtained: Figure 3 and Figure 4 As shown in the results, the VMD method has a better suppression effect on mode mixing in resistivity monitoring data than the EMD method.

[0055] S2, perform FFT on the modal component set Γ1(t) obtained in S1 to obtain the frequency domain signal set Γ1(f)={imf1(f),imf2(f),…,imf6(f)}, as shown Figure 5 As shown;

[0056] Where imf1(f) represents the frequency domain signal corresponding to imf1(t), and f is the sampling frequency;

[0057] S3. In the resistivity monitoring project, the transmission frequency is set to f' = 15Hz. A filter is designed with boundaries [15 - 0.5 * 15, 15 + 0.5 * 15], where 0.5 is the filter bandwidth factor. This filter is used to filter Γ1(f) obtained in S2 at 15Hz. The filtered dataset is Γ1(f). 1_filter (f'), such as Figure 6 As shown;

[0058] S4, Obtain the filtered dataset Γ 1_filter The maximum amplitude of each component signal in (f') is used to obtain the component signal amplitude set Γ. 1_filter_max(f') = {0.152867, 0.118226, 0.264163, 0.766915, 0.205501, 0.148922}, compare Γ 1_filter_max The values ​​of each element in (f') are used to obtain the component corresponding to the maximum value, which is the 4th modal component, 0.766915.

[0059] S5. Since resistivity monitoring uses 15Hz AC transmission, and downhole noise interference is usually high-frequency interference, the components corresponding to the high-frequency components {imf1(t),imf2(t),imf3(t)} are removed from the modal component set Γ1(t) using the fourth modal component imf4(t) as the boundary, resulting in the non-high-frequency modal component set Γ. 1_delete (t)={imf4(t),imf5(t),imf6(t)};

[0060] S6, the non-high frequency modal component set Γ obtained in S5 1_delete The elements in (t) are added together to obtain the reconstructed mine time-shift resistivity monitoring signal Γ1_re(t)={imf4(t)+imf5(t)+imf6(t)}, as shown below. Figure 7 As shown;

[0061] S7. Perform an FFT on Γ1_re(t) obtained in S6 to obtain the corresponding frequency domain signal Γ. 1_re (f'), and extract the amplitude of the Γ1_re(f') signal at a frequency of 15Hz, denoted as Γ. i_re_amp =0.86V;

[0062] S8, the resistivity ρ1 of the formation at the first measuring point of the resistivity monitoring project site is 1000 Ω·m. Record the magnitude of the current I generated during the measurement. i The current is 0.1A, and the position of the straight-line distance between the observation point and the excitation point in the original measurement data is recorded as r1 = 10m. The potential is calculated according to the formula in the DC field. The analytical solution voltage at the location corresponding to the first measurement data is obtained and denoted as U1 = 0.80V;

[0063] Here, the excitation point represents the point where the signal is emitted;

[0064] S9, repeat S1-S8 on the data from the remaining 9 measurement points to obtain the set Γ of amplitude values ​​at 15Hz for the 10 measurement points. all _re_amp = {0.86, 0.78, 0.63, 0.54, 0.48, 0.42, 0.38, 0.36, 0.33, 0.29}, as shown Figure 8As shown, and the set of analytical voltages at 10 measuring points, Ual = {0.80, 0.66, 0.57, 0.50, 0.44, 0.40, 0.36, 0.33, 0.31, 0.28}, as... Figure 9 As shown;

[0065] S10, calculate the amplitude percentage at the location corresponding to the first measurement point at a frequency of 15Hz. The percentage of analytical voltage is

[0066] For the data from the first measurement, the formula is used. The corrected voltage signal is calculated by applying an analytical solution to the data from the first measurement to correct the measured data, resulting in the corrected signal U. out_1 =0.79V;

[0067] S11, repeat S10 on the data from the remaining 9 measuring points to obtain the noise-reduced mine time-shift resistivity monitoring signal. This refers to the noise-reduced time-shift resistivity monitoring signal from the mine, such as... Figure 10 As shown.

[0068] Figure 11 The graph shows a comparison of signal errors after voltage trend correction. As can be seen from the graph, the signal error after voltage trend correction is significantly higher than the signal error without voltage trend correction.

[0069] The preferred embodiments of this disclosure have been described in detail above with reference to the accompanying drawings. However, this disclosure is not limited to the specific details of the above embodiments. Within the scope of the technical concept of this disclosure, various simple modifications can be made to the technical solutions of this disclosure, and these simple modifications all fall within the protection scope of this disclosure.

[0070] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. In order to avoid unnecessary repetition, this disclosure will not describe the various possible combinations separately.

[0071] Furthermore, the various implementation methods disclosed in this solution can be combined in any way, as long as they do not violate the spirit of this disclosure, they should also be regarded as the content invented by this disclosure.

Claims

1. A method for noise reduction of mine time-lapse resistivity monitoring data, characterized in that, Includes the following steps, S0, acquire m full waveform signals data(t) from m measurement points; S1, for the i-th full waveform signal data i (t) Perform VMD decomposition to obtain imf of k modal components. k The modal component set composed of (t) Γ i (t)={imf1(t),imf2(t),…,imf k (t)}; Where t represents the sampling time, the measurement point is the observation point, and the i is initially 1; S2, the set of modal components Γ obtained from S1 i (t) Perform FFT to obtain the frequency domain signal set. G i (f)={imf1(f),imf2(f),…,imf k (f)}; Among them, imf k (f) represents imf k (t) represents the frequency domain signal, where f is the sampling frequency; S3, setting the transmission frequency as f' in the resistivity monitoring project, and processing the frequency domain signal set Γ obtained in S2. i (f) Filtering is performed at the transmission frequency f' to obtain the filtered dataset Γ. i_filter (f'), the filtered dataset Γ i_filter (f') is the set of k component signals; S4, Obtain the filtered dataset Γ i_filter The maximum amplitude of each component signal in (f') is used to obtain the component signal amplitude set. Γ i_filter_max (f')={imf 1_filter_max (f'),imf 2_filter_max (f'),…,imf k_filter_max (f')}, compare Γ i_filter_max The size of each element in (f') is used to obtain the component corresponding to the maximum value, which is the j-th modal component imf. j (t); Among them, imf k_filter_max (f') represents the filtered dataset Γ i_filter The maximum amplitude value corresponding to the kth component signal in (f'); S5, with the j-th modal component imf j (t) is the boundary, in the modal component set Γ i Remove the components corresponding to high-frequency components from (t) {imf1(t),imf2(t),...,imf j-1 (t)}, thus obtaining the non-high frequency modal component set Γ i_delete (t)={imf j (t),imf j+1 (t),…,imf k (t)}; S6, the non-high frequency modal component set Γ obtained in S5 i_delete Adding the elements in (t) yields the reconstructed mine time-shift resistivity monitoring signal Γ. i_re (t); S7, Γ obtained from S6 i_re (t) is subjected to FFT to obtain the corresponding frequency domain signal Γ. i_re (f'), and extract Γ i_re The amplitude of the signal at frequency f' is denoted as Γ. i_re_amp ; S8, obtain the analytical voltage U at the i-th measurement point. i ; S9, let i = i + 1, repeat S1-S8 on the data of the remaining m-1 measurement points to obtain the set Γ of the amplitudes of the m measurement points at frequency f'. all_re_amp ={Γ 1_re_amp ,Γ 2_re_amp ,…,Γ m_re_amp }, and the set U of analytical voltages at m measurement points. all ={U1,U2,…,U m }; S10, the corrected voltage signal U of the i-th measuring point is obtained through equation (1). out_i ; Among them, Γ re_i U represents the amplitude percentage of the i-th measurement point at frequency f', as shown in equation (2); the initial value of i is 1; U re_i The percentage of the analytical voltage at the i-th measurement point at frequency f' is represented by equation (3); S11, let i = i + 1, repeat S10 for the data of the remaining m-1 measuring points to obtain the noise-reduced mine time-shift resistivity monitoring signal U. out_all , 2. The method for noise reduction of mine time-lapse resistivity monitoring data as described in claim 1, characterized in that, In S3, the frequency domain signal set Γ is filtered. i (f) Filtering is performed, and the boundary of the filter is [f'-αf', f'+αf']; where α is the filter bandwidth factor.

3. The method for noise reduction of mine time-lapse resistivity monitoring data as described in claim 2, characterized in that, The value of α is 0.

5.

4. The method for noise reduction of mine time-shift resistivity monitoring data as described in any one of claims 2 or 3, characterized in that, The analytical voltage in S8 Where, ρ i I represents the formation resistivity at the i-th measuring point. i r represents the excitation current at the i-th measuring point during measurement. i This represents the straight-line distance between the observation point and the excitation point in the original measurement data, where the excitation point represents the point where the signal is transmitted.

Citation Information

Patent Citations

  • Microseismic signal noise reduction filtering method based on VMD and wavelet packet

    CN107515424A

  • High-voltage cable partial discharge on-line monitoring method and system

    CN114089138A