Pipeline leak multi-leak point identification and positioning method
By employing methods such as generalized cross-correlation analysis, wavelet analysis, and Hilbert envelope analysis, combined with the sliding window method and time delay estimation, the problem of identifying and locating multiple leaks was solved, improving the efficiency and accuracy of pipeline leak detection and preventing missed detections.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- ANHUI ZHIBO PHOTOELECTRIC TECHNOLOGY CO LTD
- Filing Date
- 2023-12-26
- Publication Date
- 2026-07-03
AI Technical Summary
Existing technologies are insufficient to effectively identify and locate multiple pipeline leaks, especially in complex working environments, leading to serious underreporting problems.
By employing generalized cross-correlation analysis, wavelet analysis, Hilbert envelope analysis, and peak detection methods, combined with the sliding window method and time delay estimation, multiple leakage points are identified and located by setting threshold judgments and distance conditions.
It enables accurate identification and location of multiple leak points, improves the efficiency and accuracy of pipeline leak detection, prevents missed detections, and provides intuitive and concise results.
Smart Images
Figure CN117743901B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of fiber optic sensing technology for pipeline leaks, and in particular to a method for identifying and locating multiple leak points in a pipeline. Background Technology
[0002] In industrial production, petrochemicals, water utilities, and other fields, pipeline systems serve as indispensable infrastructure, undertaking the crucial task of transporting various liquids and gases. These pipeline networks, like lifelines, run through the entire industrial system, ensuring the efficient flow of resources. However, due to long-term operation, harsh working environments, and constantly changing operating conditions, pipeline systems face numerous challenges, the most prominent being pipeline leakage. Pipeline leaks not only lead to resource waste but can also cause environmental pollution and safety accidents, severely impacting a company's economic benefits and social reputation. Traditional pipeline leak detection methods primarily focus on identifying and locating single leak points. While effective in specific situations, these methods often cannot handle more complex multi-leak situations.
[0003] Currently, correlation analysis is a widely used method in pipeline leak detection. This method involves placing sensors at both ends of the leak point to simultaneously collect acoustic signals, and then identifying and locating the leak point by analyzing the correlation between the two acoustic signals. However, research on multiple leak points is scarce. In complex real-world operating environments, a pipeline may have multiple leak points. Current technologies cannot effectively address this situation, leading to missed leaks.
[0004] To address the aforementioned issues, this invention extracts peak values using a sliding window method and filters data by setting distance conditions to ensure intervals between peak values. Each peak value is then thresholded to identify multiple leaks. Furthermore, time delay estimation accurately determines the location of each leak, effectively solving the problem of identifying and locating multiple leaks, improving the efficiency and accuracy of pipeline leak detection, and preventing missed detections. In addition, the method and results are presented intuitively and concisely, enabling rapid determination of the number and location of multiple pipeline leaks. Summary of the Invention
[0005] To address the aforementioned problems, this invention provides a method for identifying and locating multiple leaks in pipelines. This invention utilizes generalized cross-correlation analysis, wavelet analysis, Siberian envelope analysis, and peak detection to identify multiple leaks and accurately determine their locations through time delay estimation. This effectively solves the problem of identifying and locating multiple leaks, improving the efficiency and accuracy of pipeline leak detection and preventing false negatives. Furthermore, the method and results are presented intuitively and concisely, enabling rapid determination of the number of leaks and their location.
[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows.
[0007] A method for identifying and locating multiple leak points in a pipeline includes the following steps:
[0008] Step 1: Install two sensors at both ends of the pipe at the suspected leak point. The two sensors will simultaneously collect the sound signals generated by the pipe leak.
[0009] Step 2: Calculate the generalized cross-correlation function of the acoustic signals received by the two sensors;
[0010] Step 3: Perform wavelet decomposition and thresholding on the generalized cross-correlation function obtained in Step 2, and reconstruct the reconstructed signal by taking the wavelet coefficients after thresholding.
[0011] Step 4: Calculate the Hilbert envelope based on the signal obtained in Step 3;
[0012] Step 5: Perform peak detection on the envelope obtained in Step 4;
[0013] Step 6: Perform threshold judgment on the peak value obtained in Step 5. If the set threshold condition is met, it is determined to be a leakage point.
[0014] Step 7: Estimate the time delay for each leak point and calculate the location of each leak point.
[0015] Preferably, the specific implementation of the generalized cross-correlation function is as follows:
[0016] Step S1: The signals received by the two sensors are respectively x 1( t )and x 2( t ),right x 1( t )and x 2( t Perform a Fourier transform to obtain X 1( ω )and X 2( ω );
[0017] Step S2: Calculate the cross-power spectral density of the two received signals. S x1x2 ( ω ):
[0018]
[0019] Where * denotes complex conjugation;
[0020] Step S3: Calculate the cross-power spectral density S x1x2( ω By performing an inverse Fourier transform, the generalized cross-correlation function of the two received signals can be obtained. C x1x2 ( t ):
[0021]
[0022] Preferably, the wavelet decomposition uses the sym4 wavelet basis function.
[0023] Preferably, the wavelet decomposition steps are as follows:
[0024] Step S1: For the generalized cross-correlation function C x1x2 ( t Perform 8-level wavelet decomposition;
[0025] Step S2: Use the hard thresholding method to perform thresholding on the wavelet coefficients of each decomposed layer to remove noise components;
[0026] Step S3: Reconstruct the wavelet coefficients after thresholding to obtain the reconstructed signal. D x1x2 ( t ).
[0027] Preferably, the hard thresholding method for removing noise components is as follows:
[0028]
[0029] in, The wavelet coefficients after decomposition The threshold is set to 0 when the wavelet coefficients are less than the threshold, while the wavelet coefficients greater than the threshold remain unchanged.
[0030] Preferably, the calculation steps for the Hilbert envelope are as follows:
[0031] Step S1: Perform Hilbert transform on the signal after wavelet analysis in step 3:
[0032]
[0033] Step S2: Calculate the envelope:
[0034]
[0035] Preferably, the peak detection steps are as follows:
[0036] Step S1: Select a window of appropriate size as the sliding window, and place the sliding window at the starting position of the envelope obtained in step 4;
[0037] Step S2: Within the sliding window, traverse the data and find the maximum value, which is the peak value;
[0038] Step S3: Move the sliding window one element to the right and continue searching for new peaks within the new sliding window;
[0039] Step S4: Repeat the above steps until the right boundary of the sliding window reaches the end of the envelope;
[0040] Step S5: Filter the data by setting distance conditions to ensure that there is a certain interval between each peak;
[0041] Step S6: Return the peak positions filtered by distance criteria.
[0042] Preferably, the appropriate window size is 5ms; the distance condition is that when adjacent peaks are ≤2ms, the largest peak is retained.
[0043] Preferably, the threshold in step 6 is set to be greater than 0.2.
[0044] Preferably, the delay estimation steps are as follows:
[0045] Step S1: Obtain the time delay corresponding to the peak value of each leakage point in Step 5. ;
[0046] Step S2: Calculate the distance of each leak point relative to the first sensor using the formula:
[0047]
[0048] Where L is the distance between the two sensors, and v is the speed at which the leakage sound propagates.
[0049] By adopting the above technical solution, the present invention has the following beneficial effects.
[0050] (1) This invention uses pipeline leakage point identification and location methods such as generalized cross-correlation analysis, wavelet analysis, Siberian envelope analysis and peak detection to identify multiple leak points and accurately obtain the location of each leak point through time delay estimation. This can effectively solve the problem of identifying and locating multiple leak points, improve the efficiency and accuracy of pipeline leakage detection, and prevent false alarms.
[0051] (2) The present invention extracts the peak value by sliding window method and filters the data by setting distance conditions to ensure that there is an interval between each peak value. Then, the peak value is judged by threshold, which can accurately determine multiple pipeline leakage points and prevent missed reports.
[0052] (3) This invention uses basic signal analysis methods. The method itself and the results are presented in an intuitive, concise and innovative way, which effectively solves the problem of identifying and locating multiple leaks and improves the efficiency and accuracy of pipeline leak detection. Attached Figure Description
[0053] The following provides a detailed discussion of the manufacture and application of preferred embodiments of the present invention. However, it should be understood that the present invention provides many applicable inventive concepts that can be embodied in various specific environments. The specific embodiments discussed are merely illustrative of specific ways of manufacturing and using the present invention and do not limit the scope of the invention. For those skilled in the art, other drawings can be obtained from these drawings without any creative effort.
[0054] Figure 1 This is a schematic diagram of the method flow of the present invention.
[0055] Figure 2 This is a schematic diagram of the two-leak pipeline leakage detection model of the present invention.
[0056] Figure 3 This is a diagram showing the leak detection results of the present invention. Detailed Implementation
[0057] The following provides a detailed discussion of the manufacture and application of preferred embodiments of the present invention. However, it should be understood that the present invention provides many applicable inventive concepts that can be embodied in various specific environments. The specific embodiments discussed are merely illustrative of specific ways of manufacturing and using the invention and do not limit the scope of the invention.
[0058] A method for identifying and locating multiple leak points in a pipeline includes the following steps.
[0059] Step 1: Install two sensors at both ends of the pipe at the suspected leak point. The two sensors simultaneously collect the sound signals generated by the pipe leak.
[0060] Step 2: Calculate the generalized cross-correlation function of the acoustic signals received by the two sensors.
[0061] Step 3: Perform wavelet decomposition and thresholding on the generalized cross-correlation function obtained in Step 2, and reconstruct the reconstructed signal by taking the wavelet coefficients after thresholding.
[0062] Step 4: Calculate the Hilbert envelope based on the signal obtained in Step 3.
[0063] Step 5: Perform peak detection on the envelope obtained in Step 4.
[0064] Step 6: Perform threshold judgment on the peak value obtained in Step 5. If the set threshold condition is met, it is determined to be a leakage point.
[0065] Step 7: Estimate the time delay for each leak point and calculate the location of each leak point.
[0066] The specific implementation of the generalized cross-correlation function is as follows.
[0067] Step S1: The signals received by the two sensors are respectively x 1( t )and x 2( t ),right x 1( t )and x 2( t Perform a Fourier transform to obtain X 1( ω )and X 2( ω ).
[0068] Step S2: Calculate the cross-power spectral density of the two received signals. S x1x2 ( ω ):
[0069]
[0070] Where * denotes complex conjugation.
[0071] Step S3: Calculate the cross-power spectral density S x1x2 ( ω By performing an inverse Fourier transform, the generalized cross-correlation function of the two received signals can be obtained. C x1x2 ( t ):
[0072] .
[0073] The wavelet decomposition uses the sym4 wavelet basis function. The steps of the wavelet decomposition are as follows.
[0074] Step S1: For the generalized cross-correlation function C x1x2 ( t Perform 8-level wavelet decomposition.
[0075] Step S2: Use the hard thresholding method to perform thresholding on the wavelet coefficients of each decomposed layer to remove noise components.
[0076] Step S3: Reconstruct the wavelet coefficients after thresholding to obtain the reconstructed signal. D x1x2 ( t ).
[0077] The hard thresholding method for removing noise components is as follows.
[0078]
[0079] in, The wavelet coefficients after decomposition The threshold is set to 0 when the wavelet coefficients are less than the threshold, while the wavelet coefficients greater than the threshold remain unchanged.
[0080] The calculation steps for the Hilbert envelope are as follows.
[0081] Step S1: Perform Hilbert transform on the signal after wavelet analysis in step 3:
[0082] .
[0083] Step S2: Calculate the envelope:
[0084] .
[0085] The peak detection steps are as follows.
[0086] Step S1: Select a window of appropriate size as the sliding window, and place the sliding window at the starting position of the envelope obtained in step 4.
[0087] Step S2: Within the sliding window, traverse the data and find the maximum value, which is the peak value.
[0088] Step S3: Move the sliding window one element to the right and continue searching for new peaks within the new sliding window.
[0089] Step S4: Repeat the above steps until the right boundary of the sliding window reaches the end of the envelope.
[0090] Step S5: Filter the data by setting distance conditions to ensure that there is a certain interval between each peak.
[0091] Step S6: Return the peak positions filtered by distance criteria.
[0092] The appropriate window size is 5ms. The distance condition is that when adjacent peaks are ≤2ms, the largest peak is retained. The threshold in step 6 is set to be greater than 0.2.
[0093] The delay estimation steps are as follows:
[0094] Step S1: Obtain the time delay corresponding to the peak value of each leakage point in Step 5. .
[0095] Step S2: Calculate the distance of each leak point relative to the first sensor using the formula:
[0096]
[0097] Where L is the distance between the two sensors, and v is the speed at which the leakage sound propagates.
[0098] The following is in conjunction with the appendix Figure 1-3 Provide a detailed description.
[0099] like Figure 1 The diagram shows a flowchart of a method for identifying and locating multiple leak points in a pipeline. It includes the following steps.
[0100] Step 1: Install two sensors at both ends of the pipe at the suspected leak point. The two sensors synchronously collect the acoustic signals generated by the leak. In this embodiment, there are two leak points, such as... Figure 2 As shown.
[0101] Step 2: Calculate the generalized cross-correlation function of the acoustic signals received by the two sensors. The generalized cross-correlation function is calculated in the following way.
[0102] Step S21: The signals received by the two sensors are respectively x 1( t )and x 2( t ),right x 1( t )and x 2( t Perform a Fourier transform to obtain X 1( ω )and X 2( ω ).
[0103] Step S22: Calculate the cross-power spectral density of the two received signals. S x1x2 ( ω ):
[0104]
[0105] Where * denotes complex conjugation.
[0106] Step S23: Calculate the cross-power spectral density S x1x2 ( ω By performing an inverse Fourier transform, the generalized cross-correlation function of the two received signals can be obtained. C x1x2 ( t ):
[0107] .
[0108] Step 3: Perform wavelet decomposition and thresholding on the generalized cross-correlation function obtained in Step 2. The wavelet decomposition uses the sym4 wavelet basis function. The steps of the wavelet decomposition are as follows.
[0109] Step S31: For the generalized cross-correlation function C x1x2 ( t Perform 8-level wavelet decomposition.
[0110] Step S32: The wavelet coefficients of each decomposed layer are thresholded using a hard thresholding method to remove noise components; the method of removing noise components using the hard thresholding method is as follows:
[0111]
[0112] in, The wavelet coefficients after decomposition The threshold is set to 0 when the wavelet coefficients are less than the threshold, while the wavelet coefficients greater than the threshold remain unchanged.
[0113] Step S33: Reconstruct the wavelet coefficients after thresholding to obtain the reconstructed signal. D x1x2 ( t ).
[0114] Step 4: Calculate the Hilbert envelope based on the signal obtained in Step 3; the calculation steps for the Hilbert envelope are as follows.
[0115] First, perform a Hilbert transform on the signal after wavelet analysis in step 3:
[0116]
[0117] The envelope is then calculated:
[0118] .
[0119] Step 5: Perform peak detection on the envelope obtained in Step 4; the peak detection steps are as follows.
[0120] Step S51: Select a window of size 5ms as the sliding window and place the sliding window at the starting position of the envelope obtained in step 4.
[0121] Step S52: Within the sliding window, traverse the data and find the maximum value, which is the peak value.
[0122] Step S53: Move the sliding window one element to the right and continue searching for new peaks within the new sliding window.
[0123] Step S54: Repeat the above steps until the right boundary of the sliding window reaches the end of the envelope.
[0124] Step S55: Data filtering is performed by setting distance conditions to ensure that there is a certain interval between each peak; the distance condition is that when adjacent peaks are ≤2ms, the largest peak is retained.
[0125] Step S56: Return the peak positions filtered by distance criteria.
[0126] Step 6: Perform a threshold judgment on the peak value obtained in Step 5. If the threshold condition of threshold > 0.2 is met, it is determined to be a leakage point.
[0127] Step 7: Estimate the time delay for each leak point and calculate the location of each leak point. The steps for time delay estimation are as follows.
[0128] First, obtain the time delay corresponding to the peak value of each leak point in step 5. Then, the distances of each leak point relative to the first sensor are calculated using the formula:
[0129]
[0130] Where L is the distance between the two sensors, and v is the speed at which the leakage sound propagates.
[0131] like Figure 3 As shown, in this embodiment, L=95m, the sound propagation speed of leakage v=1250m / s, and the time delay corresponding to the peak values at the two leakage points. =63ms and 43ms, can be found Figure 3 It is clear that two points have peak values exceeding the threshold of 0.2, indicating that there are two missing points. Based on the time delay estimation, the time delay of the two peak points is... Substituting 63ms and 42ms into the formula: We can calculate that the distance between leak point 1 and sensor is 8.12 meters, and the distance between leak point 2 and sensor 1 is 21.25 meters.
[0132] In summary, the leak identification result diagram obtained based on this method can intuitively reflect the number of leaks in the pipeline by the number of peaks exceeding the judgment threshold; through... Figure 3 By substituting the time delay data of the two leak points on the horizontal axis into the time delay estimation formula, the location of the corresponding leak point can be accurately calculated. This method is intuitive and simple, and it is accurate in detecting and locating multiple leak points in pipelines. It can significantly improve the efficiency and accuracy of pipeline leak detection, overcome the shortcomings of existing pipeline leak detection methods that cannot detect and locate multiple leak points, and prevent false negatives.
[0133] This invention extracts peak values using a sliding window method and filters data by setting distance conditions to ensure intervals between peak values. Each peak value is then thresholded to identify multiple leaks. By estimating time delay, the location of each leak is accurately determined, effectively solving the problem of identifying and locating multiple leaks and improving the efficiency and accuracy of pipeline leak detection. Furthermore, the method and results are presented intuitively and concisely, enabling rapid determination of the number and location of multiple pipeline leaks.
[0134] Although the specification has provided a detailed description, it should be understood that various changes, substitutions, and modifications can be made without departing from the spirit and scope of the invention as defined by the appended claims. Furthermore, the specific embodiments described are not intended to limit the scope of the invention, and those skilled in the art will readily understand based on this invention that existing or future-developed processes, machines, manufactures, compositions of matter, means, methods, or steps can perform substantially the same functions or achieve substantially the same results as the embodiments of the invention. Therefore, the appended claims are intended to include such processes, machines, manufactures, compositions of matter, means, methods, or steps within their scope.
Claims
1. A method for identifying and locating multiple leak points in a pipeline, characterized in that: Includes the following steps: S1: Install two sensors at both ends of the pipe at the suspected leak point. The two sensors will simultaneously collect the acoustic signal generated by the pipe leak. S2: Calculate the generalized cross-correlation function of the acoustic signals received by the two sensors; the specific implementation of the generalized cross-correlation function is as follows: S21: The signals received by the two sensors are respectively x 1( t )and x 2( t ),right x 1( t )and x 2( t Perform a Fourier transform to obtain X 1( ω )and X 2( ω ); S22: Calculate the cross power spectral density of the two received signals. S x1x2 ( ω ): in, Indicates complex conjugation; S23: Cross-power spectral density S x1x2 ( ω By performing an inverse Fourier transform, the generalized cross-correlation function of the two received signals can be obtained. C x1x2 ( t ): ； S3: Perform wavelet decomposition and thresholding on the generalized cross-correlation function obtained in S2, and reconstruct the reconstructed signal using the wavelet coefficients after thresholding. The wavelet decomposition uses the sym4 wavelet basis function. The steps of the wavelet decomposition are as follows: S31: For the generalized cross-correlation function C x1x2 ( t Perform 8-level wavelet decomposition; S32: The wavelet coefficients of each decomposed layer are thresholded using a hard thresholding method to remove noise components; the method of removing noise components using the hard thresholding method is as follows: in, The wavelet coefficients after decomposition The threshold is set to 0 when the wavelet coefficients are less than the threshold, while the wavelet coefficients greater than the threshold remain unchanged. S33: Reconstruct the wavelet coefficients after thresholding to obtain the reconstructed signal. D x1x2 ( t ); S4: Calculate the Hilbert envelope based on the signal obtained in S3; the calculation steps for the Hilbert envelope are as follows: S41: Perform Hilbert transform on the signal processed by wavelet analysis in S3: S42: Calculate the envelope: ； S5: Perform peak detection on the envelope obtained in S4; the peak detection steps are as follows: S51: Select a window of appropriate size as a sliding window, and place the sliding window at the starting position of the envelope obtained in S4; S52: Within the sliding window, traverse the data and find the maximum value, which is the peak value; S53: Move the sliding window one element to the right and continue searching for new peaks within the new sliding window; S54: Repeat the above steps until the right boundary of the sliding window reaches the end of the envelope; S55: Data filtering is performed by setting distance conditions to ensure that a certain interval is maintained between peak values; S56: Returns the peak position filtered by distance criteria; S6: Perform threshold judgment on the peak value obtained in S5. If the set threshold condition is met, it is determined to be a leakage point. S7: Estimate the time delay for each leak point and calculate the location of each leak point.
2. The method for identifying and locating multiple leak points in a pipeline as described in claim 1, characterized in that: The appropriate window size is 5ms; the distance condition is that when adjacent peaks are ≤2ms, the largest peak is retained.
3. The method for identifying and locating multiple leak points in a pipeline as described in claim 1, characterized in that: The threshold in S6 is set to be greater than 0.
2.
4. The method for identifying and locating multiple leak points in a pipeline as described in claim 1, characterized in that: The delay estimation steps are as follows: S71: Obtain the time delay corresponding to the peak value of each leak point in S5. ; S72: The distances of each leak point relative to the first sensor are calculated using a formula: Where L is the distance between the two sensors, and v is the speed at which the leakage sound propagates.