Underwater oil and gas equipment leakage three-dimensional sound source positioning algorithm based on hydrophone

Abstract

The present application belongs to the field of petroleum engineering, and particularly relates to a three-dimensional sound source positioning algorithm for underwater oil and gas equipment leakage based on a hydrophone. The three-dimensional sound source positioning algorithm for underwater oil and gas equipment leakage based on the hydrophone comprises three steps: parameter optimization of variational mode decomposition based on grey wolf optimization algorithm, signal reconstruction based on variational mode decomposition, and leakage positioning based on time difference of arrival. The parameter optimization of variational mode decomposition based on the grey wolf optimization algorithm can obtain the optimal values of the decomposition layer number K and the penalty factor alpha for variational mode decomposition. The signal reconstruction based on variational mode decomposition is to select the signal components with high correlation to the original signals of the underwater oil and gas equipment leakage hydrophone according to the threshold value and complete reconstruction. The leakage positioning based on the time difference of arrival is to calculate the time difference of the leakage signal arriving at each hydrophone according to the reconstructed signal, and finally obtain the three-dimensional coordinates of the leakage source by substituting the time difference of arrival into the arrival time difference positioning equation.

Description

Technical Field

[0001] This invention belongs to the field of petroleum engineering, specifically, it relates to a three-dimensional sound source localization algorithm for underwater oil and gas equipment leakage based on hydrophones. Background Technology

[0002] Leaks in underwater oil and gas equipment not only affect the normal production of offshore oil fields and threaten the lives of workers, but also cause marine environmental pollution, trigger ecological disasters, and seriously hinder the development of offshore oil and gas development. Therefore, it is particularly important to detect leaks and accurately locate the leak points.

[0003] In response to the current demand for underwater leak detection in the marine field, hydrophones are indispensable equipment in underwater acoustic measurement. Certain models of hydrophones can operate normally at depths of 1000 meters and withstand hydrostatic pressures up to 100 bar. Furthermore, they are corrosion-resistant and can tolerate high salinity. Therefore, hydrophone-based leak detection is one of the most direct, accurate, and reliable solutions for leak detection and identification. However, hydrophones are best suited for low signal-to-noise ratio environments. The marine environment is often complex and variable due to the influence of wind, waves, ocean currents, and marine biological activity, and underwater sound signals are often mixed with various noises. Therefore, developing a three-dimensional sound source localization algorithm for underwater oil and gas equipment leaks based on hydrophones is of great significance. Summary of the Invention

[0004] To overcome the shortcomings of existing technologies, this invention provides a three-dimensional sound source localization algorithm for underwater oil and gas equipment leaks based on hydrophones.

[0005] To achieve the above objectives, such as Figure 1 As shown, the three-dimensional sound source localization algorithm for underwater oil and gas equipment leaks based on hydrophones includes three steps: variational mode decomposition parameter optimization based on the Grey Wolf optimization algorithm, signal reconstruction based on variational mode decomposition, and leak localization based on time difference of arrival.

[0006] The specific steps for variational mode decomposition parameter optimization based on the Grey Wolf optimization algorithm are as follows:

[0007] S101: Initialize the gray wolf population and calculate the average envelope entropy;

[0008] S102: Initialize the iteration number t=1, initialize the decreasing component a, the convergent control coefficient vector A and the random control coefficient vector C;

[0009] S103: Calculate the fitness of individual gray wolves and save the top three wolves with the best current fitness (W). α W β and W δ ;

[0010] S104: Update the current position of the gray wolf, update the decreasing component a, the convergent control coefficient vector A and the random control coefficient vector C;

[0011] S105: Calculate the fitness of all gray wolves and update the top three wolves with the best current fitness (W). α W β and W δ The fitness and location.

[0012] The specific steps for signal reconstruction based on variational mode decomposition are as follows:

[0013] S201: Based on the optimal values ​​of the number of decomposition layers K and the penalty factor α obtained from the Grey Wolf optimization algorithm, variational mode decomposition is performed on the original signal of the underwater oil and gas equipment leakage hydrophone.

[0014] S202: By performing variational mode decomposition on the original signal of the underwater oil and gas equipment leakage hydrophone, K characteristic mode functions with fixed bandwidth and center frequency are obtained;

[0015] S203: Calculate the cross-correlation coefficient between each characteristic mode function and the original signal of the underwater oil and gas equipment leakage hydrophone;

[0016] S204: Select signal components that are highly correlated with the original signal of the underwater oil and gas equipment leakage hydrophone according to the threshold.

[0017] S205: Reconstruct the original signal of the underwater oil and gas equipment leakage hydrophone to achieve noise reduction of the leakage signal.

[0018] The specific steps for leak location based on time difference of arrival are as follows:

[0019] S301: Calculate the sampling point difference based on the reconstructed signal;

[0020] S302: Based on the sampling point difference of the same sudden peak of the noise reduction signal of each channel hydrophone, obtain the time difference between the arrival of the leakage signal at each hydrophone, and calculate the time difference based on the obtained sampling point difference;

[0021] S303: Establish a location equation based on the time difference of arrival (TDOA) of the hydrophone placement position, solve the location equation set using the two-stage weighted least squares method, and calculate the three-dimensional coordinates of the leakage source.

[0022] S304: The accuracy of underwater leak location is an important indicator used to evaluate whether the leak location is accurate. Its accuracy is expressed by the root mean square error. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of a three-dimensional sound source localization algorithm for underwater oil and gas equipment leaks based on hydrophones. Detailed Implementation

[0024] like Figure 1 As shown, the three-dimensional sound source localization algorithm for underwater oil and gas equipment leaks based on hydrophones includes three steps: variational mode decomposition parameter optimization based on the Grey Wolf optimization algorithm, signal reconstruction based on variational mode decomposition, and leak localization based on time difference of arrival.

[0025] The specific steps for variational mode decomposition parameter optimization based on the Grey Wolf optimization algorithm are as follows:

[0026] S101: Initialize the gray wolf population and calculate the average envelope entropy value. The average envelope entropy refers to the average value of the envelope entropy of each modal component obtained by variational mode decomposition of the signal under specific parameters K and α. The (K, α) parameter pair is used as the position vector that the gray wolf population needs to optimize in the gray wolf optimization algorithm.

[0027] S102: Initialize the iteration number t=1, initialize the decreasing component a, the convergent control coefficient vector A, and the random control coefficient vector C. When wolves hunt, they first surround the prey. The system of equations for their surrounding behavior is as follows:

[0028]

[0029]

[0030] Where t represents the current iteration number, X p Let X be the position vectors of the prey and the gray wolf, and let A be the convergent control coefficient vector and C be the random control coefficient vector, calculated using the following equation:

[0031]

[0032]

[0033] r1 and r2 are random numbers between 0 and 1, and the decreasing component a decreases linearly from 2 to 0 during the iteration.

[0034] S103: Calculate the fitness of individual gray wolves and save the top three wolves with the best current fitness (W). α W β and W δ Save the three best solutions obtained up to the current iteration number, and other search agents update their own positions based on the position of the best search agent;

[0035] S104: Update the current position of the gray wolf, update the decreasing component a, the convergent control coefficient vector A and the random control coefficient vector C. In the process of constructing the attack prey model, A is a random vector in the interval [-a, a], where a decreases linearly during the iteration process. When A is in the interval [-1, 1], the next position of the search agent is at a random position between the current gray wolf and the prey.

[0036] S105: Calculate the fitness of all gray wolves and update the top three wolves with the best current fitness (W). α W β and W δ If the fitness and position of the feature mode function are reached, the optimal parameter pair (K, α) is output if the maximum number of iterations mI is reached, where K is the number of decomposition layers of the feature mode function and α is the penalty factor.

[0037] The specific steps for signal reconstruction based on variational mode decomposition are as follows:

[0038] S201: Based on the optimal values ​​of the number of decomposition layers K and the penalty factor α obtained from the Grey Wolf optimization algorithm, variational mode decomposition is performed on the original signal of the underwater oil and gas equipment leakage hydrophone.

[0039] S202: By performing variational mode decomposition on the original signal of the underwater oil and gas equipment leakage hydrophone, K characteristic mode functions with fixed bandwidth and center frequency are obtained;

[0040] S203: Calculate the cross-correlation coefficient between each characteristic mode function and the original signal from the underwater oil and gas equipment leak hydrophone. The magnitude of the cross-correlation coefficient reflects the correlation between the two variables. After variational mode decomposition, a large cross-correlation coefficient between a certain mode component and the original signal indicates that the mode component is highly correlated with the original signal. The cross-correlation coefficient is calculated as follows:

[0041]

[0042] Where R represents the cross-correlation coefficient between the characteristic mode function and the original signal from the underwater oil and gas equipment leak hydrophone, and x(t) and u i (t) represent the original signal of the underwater oil and gas equipment leak hydrophone and the characteristic mode function, respectively. E and D represent the expected value and variance in mathematics, respectively. The closer the R value is to 1, the more correlated the characteristic mode function is with the original signal of the underwater oil and gas equipment leak hydrophone, and the better it reflects the characteristics of the original signal of the underwater oil and gas equipment leak hydrophone. The closer the R value is to 0, the weaker the correlation between the characteristic mode function and the original signal of the underwater oil and gas equipment leak hydrophone, and the closer it is to the noise signal.

[0043] S204: Select the signal components that are highly correlated with the original signal of the underwater oil and gas equipment leakage hydrophone according to the threshold. The signal components that are less than the threshold are noise signal components, and the signal components that are greater than or equal to the threshold are useful signal components.

[0044] S205: Reconstruct the original signal of the underwater oil and gas equipment leakage hydrophone to achieve noise reduction of the leakage signal.

[0045] The specific steps for leak location based on time difference of arrival are as follows:

[0046] S301: Calculate the sampling point difference based on the reconstructed signal, reduce the noise of the original signal of the underwater oil and gas equipment leak hydrophone, obtain the hydrophone noise-reduced signal, and determine whether a leak has occurred by searching for the sudden peak of the hydrophone noise-reduced signal.

[0047] S302: Based on the sampling point difference of the same sudden peak of the noise reduction signal of each channel hydrophone, obtain the time difference between the arrival of the leakage signal at each hydrophone, and calculate the time difference based on the obtained sampling point difference;

[0048] S303: Establish a location equation based on the time difference of arrival (TDOA) of the hydrophone placement position, solve the location equation set using the two-stage weighted least squares method, and calculate the three-dimensional coordinates of the leakage source.

[0049] S304: The accuracy of underwater leak location is an important indicator for evaluating the accuracy of leak location. Its accuracy is expressed by the root mean square error (RMSE), and the formula for calculating the RMSE is as follows:

[0050]

[0051] Where RMSE is the root mean square error, and N represents the number of trials; v 0 v is the coordinate vector of the leak point; m Leakage point v 0 The coordinate vector estimated by the m-th positioning.

Claims

1. A three-dimensional sound source localization algorithm for underwater oil and gas equipment leaks based on hydrophones, characterized in that: It consists of three steps: variational mode decomposition parameter optimization based on the gray wolf optimization algorithm, signal reconstruction based on variational mode decomposition, and leak localization based on time difference of arrival. The specific steps for variational mode decomposition parameter optimization based on the Grey Wolf optimization algorithm are as follows: S101: Initialize the gray wolf population and calculate the average envelope entropy; S102: Initialize the iteration number t=1, initialize the decreasing component a, the convergent control coefficient vector A and the random control coefficient vector C; S103: Calculate the fitness of the gray wolf individual, save the top three wolves W with the best current fitness α , W β and W δ ; S104: Update the current position of the gray wolf, update the decreasing component a, the convergent control coefficient vector A and the random control coefficient vector C; S105: Calculate the fitness of all wolves, update the top three wolves W α , W β , and W δ 's fitness and position; The specific steps for signal reconstruction based on variational mode decomposition are as follows: S201: Based on the optimal values ​​of the number of decomposition layers K and the penalty factor α obtained from the Grey Wolf optimization algorithm, variational mode decomposition is performed on the original signal of the underwater oil and gas equipment leakage hydrophone. S202: By performing variational mode decomposition on the original signal of the underwater oil and gas equipment leakage hydrophone, K characteristic mode functions with fixed bandwidth and center frequency are obtained; S203: Calculate the cross-correlation coefficient between each characteristic mode function and the original signal of the underwater oil and gas equipment leakage hydrophone; S204: Select signal components that are highly correlated with the original signal of the underwater oil and gas equipment leakage hydrophone according to the threshold. S205: Reconstruct the original signal of the underwater oil and gas equipment leakage hydrophone to achieve noise reduction of the leakage signal; The specific steps for leak location based on time difference of arrival are as follows: S301: Calculate the sampling point difference based on the reconstructed signal; S302: Based on the sampling point difference of the same sudden peak of the noise reduction signal of each channel hydrophone, obtain the time difference between the arrival of the leakage signal at each hydrophone, and calculate the time difference based on the obtained sampling point difference; S303: Establish a location equation based on the time difference of arrival (TDOA) of the hydrophone placement position, solve the location equation set using the two-stage weighted least squares method, and calculate the three-dimensional coordinates of the leakage source. S304: The accuracy of underwater leak location is an important indicator used to evaluate whether the leak location is accurate. Its accuracy is expressed by the root mean square error.

2. The three-dimensional sound source localization algorithm for underwater oil and gas equipment leakage based on hydrophones according to claim 1, characterized in that: When wolves hunt, they first surround their prey. The equations governing their encirclement behavior are shown below: ； ； where t denotes the current iteration number, X p and X are the position vectors of the prey and the gray wolf, the convergence control coefficient vector A and the random control coefficient vector C are coefficient vectors calculated by the following equations: ； ； r1 and r2 are random numbers between 0 and 1, and the decreasing component a decreases linearly from 2 to 0 during the iteration.

3. The three-dimensional sound source localization algorithm for underwater oil and gas equipment leakage based on hydrophones according to claim 1, characterized in that: In the process of constructing the attack prey model, A is a random vector in the interval [-a, a], where a decreases linearly during the iteration process. When A is in the interval [-1, 1], the next position of the search agent is at a random position between the current gray wolf and the prey.

4. The three-dimensional sound source localization algorithm for underwater oil and gas equipment leakage based on hydrophones according to claim 1, characterized in that: The cross-correlation coefficient reflects the correlation between two variables. After variational mode decomposition, a large cross-correlation coefficient between a certain mode component and the original signal indicates a strong correlation between that mode component and the original signal. The cross-correlation coefficient is calculated as follows: ； Wherein, R represents the cross-correlation coefficient between the characteristic modal function and the original signal of the underwater oil and gas equipment leakage hydrophone, x(t) and u i (t) respectively represent the original signal of the underwater oil and gas equipment leakage hydrophone and the characteristic modal function, E and D respectively represent the expectation and variance in mathematics, and the closer the R value is to 1, the more relevant the characteristic modal function is to the original signal of the underwater oil and gas equipment leakage hydrophone, and the more the characteristic modal function can reflect the characteristics possessed by the original signal of the underwater oil and gas equipment leakage hydrophone, and the closer the value is to 0, the weaker the correlation between the characteristic modal function and the original signal of the underwater oil and gas equipment leakage hydrophone, and the closer to the noise signal.

5. The three-dimensional sound source localization algorithm for underwater oil and gas equipment leakage based on hydrophones according to claim 1, characterized in that: The formula for calculating the root mean square error is as follows: ； Where RMSE is the root mean square error, and N represents the number of trials; v 0 v is the coordinate vector of the leak point; m Leakage point v 0 The coordinate vector estimated by the m-th positioning.

Images