An underwater manifold leakage positioning method and system based on a full-electric pressure wave generator

Abstract

This invention belongs to the field of petroleum engineering, specifically relating to a method and system for locating leaks in underwater manifolds based on a fully electrically controlled pressure wave generator. The method and system include a method for locating leaks in underwater manifolds based on a fully electrically controlled pressure wave generator and a system for locating leaks in underwater manifolds based on a fully electrically controlled pressure wave generator. The method for locating leaks in underwater manifolds based on a fully electrically controlled pressure wave generator includes three steps: establishing a subsystem for the structure of the fully electrically controlled pressure wave generator, establishing a subsystem for acquiring and processing signals from underwater production manifolds, and establishing a subsystem for locating leaks in underwater production manifolds. The system for locating leaks in underwater manifolds based on a fully electrically controlled pressure wave generator includes a subsystem for the structure of the fully electrically controlled pressure wave generator, a subsystem for acquiring and processing signals from underwater production manifolds, and a subsystem for locating leaks in underwater production manifolds.

Description

Technical Field

[0001] This invention belongs to the field of petroleum engineering and relates to a method and system for locating underwater manifold leaks based on a fully electrically controlled pressure wave generator. Background Technology

[0002] Leakage in liquid environments is widespread in nature and industry, but is most common in marine environments and has significant research value in marine development. As oil and gas exploration and development gradually shifts from shallow to deep water, subsea oil and gas production systems are also transforming from the early single-satellite well development model to a cluster-satellite well development model centered on subsea manifolds. The subsea production manifold, as the core of the subsea production system, integrates structure, pipelines, valves, instruments, controls, and connections. It is mainly used to collect and transport oil and gas from the wellhead, distribute electricity and chemical reagents, and collect oil and gas from the wellhead through crossover pipes and transport it to the subsea pipeline. It is a key facility of the subsea production system.

[0003] With the large-scale development of offshore oil and gas resources, oil and gas development is gradually shifting to deep-water areas. The complexity and variability of the marine environment greatly increases the risk of leaks in underwater production manifolds. Therefore, detecting the leak source and accurately locating the leak point is crucial for addressing leaks during offshore oil and gas development. However, underwater production manifolds are located in complex environments, are large in size, and suffer from problems such as significant noise interference and the inability to detect minute leaks. Mature engineering applications and equipment for online monitoring and precise location of leaks in underwater production manifolds have not yet been developed, either domestically or internationally. There is an urgent need to develop underwater production manifold leak signal enhancement equipment, elucidate the leak detection mechanism, and develop key technologies and complete sets of equipment for underwater manifold leak accident detection. Therefore, developing an underwater manifold leak location method and system based on a fully electrically controlled pressure wave generator is of paramount importance. Summary of the Invention

[0004] To overcome the shortcomings of existing technologies, this invention provides a method and system for locating underwater manifold leaks based on a fully electrically controlled pressure wave generator.

[0005] To achieve the above objectives, according to one aspect of the present invention, an underwater manifold leakage location method and system based on a fully electrically controlled pressure wave generator includes three steps: establishing a fully electrically controlled pressure wave generator structure subsystem, establishing an underwater production manifold signal acquisition and processing subsystem, and establishing an underwater production manifold leakage location subsystem.

[0006] The specific steps for establishing the fully electrically controlled pressure wave generator subsystem are as follows:

[0007] A101: Establish a fully electronically controlled pressure wave generator transmission system;

[0008] A102: Set the pressure waveform generated by the fully electronically controlled pressure wave generator to generate a transient excitation signal;

[0009] A103: High-pressure sealing of the fully electrically controlled pressure wave generator.

[0010] The specific steps for establishing the underwater production manifold signal acquisition and processing subsystem are as follows:

[0011] A201: Establishment of an underwater production manifold signal acquisition system;

[0012] A202: Calculate the propagation velocity V of the pressure wave from a leak in an underwater production manifold under the impact of a fully electrically controlled pressure wave generator. q ;

[0013] A203: Establishment of the underwater production manifold signal processing unit. Signal processing mainly includes signal dimensionality reduction, feature extraction, signal denoising, and signal reconstruction of the acquired underwater production manifold pressure signals.

[0014] ① A signal dimensionality reduction algorithm based on improved principal component analysis is proposed. New variables are constructed using linear transformations of the extracted original feature parameters, and principal components are extracted from them for signal dimensionality reduction. The original features are projected onto the dimension with the maximum projective information content, ensuring minimal information loss after dimensionality reduction.

[0015] ② A signal processing method based on the fusion of wavelet decomposition and deep convolutional neural network is proposed, and a deep convolutional neural network model fused with wavelet decomposition is constructed. The wavelet decomposition layer and the convolutional layer are executed alternately, and the convolutional module is used for repeated feature extraction to realize feature extraction, signal denoising and signal reconstruction of the leakage pressure signal of underwater production manifold.

[0016] The specific steps for establishing the underwater production manifold leak location subsystem are as follows:

[0017] A301: Calculation of the time difference of arrival of transient pressure signals in underwater production manifolds. The reconstructed transient pressure signals of the underwater production manifolds are analyzed to extract the time difference t between the upstream and downstream signal acquisition modules that acquire the transient pressure changes. 12 ;

[0018] A302: Leak location in underwater production manifolds. Based on the pressure wave propagation velocity V of the leak in the underwater production manifold. q The distance L between the upstream signal acquisition module and the downstream signal acquisition module of the underwater production manifold, and the transient pressure wave arrival time difference t. 12 Calculate the location of the leak point to achieve leak location of underwater production manifold based on transient pressure signal enhancement.

[0019] According to another aspect of the present invention, an underwater manifold leakage location method and system based on a fully electrically controlled pressure wave generator includes a fully electrically controlled pressure wave generator structure subsystem, an underwater production manifold signal acquisition and processing subsystem, and an underwater production manifold leakage location subsystem.

[0020] The structural subsystem of the fully electrically controlled pressure wave generator includes the generator rear end cover, bearing housing, reducer, servo motor, sealing cylinder, front flange, plunger rod, load connection thread, plunger, bearing upper end cover, synchronous pulley, bearing assembly, coupling, buffer pad, guide tube, ball screw, dynamic seal ring, and underwater high-pressure sealing chamber. A servo motor connects to a reducer, providing power and speed regulation for the fully electronically controlled pressure wave generator. The servo motor and reducer are fixed by bearing housings. A synchronous pulley connects the servo motor and a coupling, sealed by the generator's rear end cover. The coupling connects to a ball screw, converting the motor's rotational motion into the screw's rotational motion. The bearing assembly is axially fixed by the upper end cover, limiting the ball screw's axial movement and ensuring rotational accuracy. The ball screw moves axially within the guide tube, driving the plunger rod in reciprocating linear motion. A buffer pad absorbs impacts and vibrations during the motion, ensuring the ball screw's accuracy and stability. The entire assembly is sealed by a sealed cylinder. The plunger rod's end has a load-bearing thread, connecting to the plunger to achieve reciprocating linear motion, generating transient excitation within the tube and emitting pressure waves. A dynamic sealing ring provides dynamic sealing during the plunger rod's reciprocating linear motion. The entire pressure wave generator structure is housed in an underwater high-pressure sealed chamber, with the front flange connected to the experimental tube flange for installation.

[0021] The underwater production manifold signal acquisition and processing subsystem includes an experimental pipe, an underwater production manifold pipeline, and an underwater production manifold signal processing unit. The experimental pipe includes an upstream signal acquisition module, a leak detection module, and a downstream signal acquisition module. The underwater production manifold pipeline includes an upstream signal acquisition module, a leak detection module, and a downstream signal acquisition module. The underwater production manifold signal processing unit includes a signal feature extraction module, a signal denoising module, and a signal reconstruction module. The front flange connects to the test pipe flange. The interior of the test pipe is polished and fits tightly with the plunger to prevent leakage. An upstream signal acquisition module, a leak detection module, and a downstream signal acquisition module are installed on the test pipe to simulate leaks and acquire pipeline signals. The test pipe connects to the underwater production manifold flange. The upstream and downstream signal acquisition modules acquire leakage signals from the leak detection module. The signals acquired by these modules are transmitted via power lines to the underwater production manifold signal processing unit. The signal processing unit includes a signal feature extraction module, a signal denoising module, and a signal reconstruction module for extracting, denoising, and reconstructing the features of the acquired raw signals. The signal processing results from the signal processing unit are transmitted via cable to the underwater production manifold leak location subsystem.

[0022] The underwater production manifold leak location subsystem includes a time difference calculation module and a leak location module. The time difference calculation module is used to calculate the time difference of arrival of the transient pressure wave of the leak signal; the leak location module is connected to the time difference calculation module via a cable to locate the leak point.

[0023] Compared to existing technologies, the beneficial effects of this invention are as follows: The fully electrically controlled pressure wave generator generates pressure waves with appropriate amplitude and frequency based on the underwater production manifold structure, pipe length, inner diameter, physical characteristics of the transported oil and gas, and internal pressure. This generates transient excitation to enhance the leakage signal, improving location accuracy and enabling leakage detection while maintaining normal transport and economic efficiency in the underwater production manifold. Based on an improved principal component analysis signal dimensionality reduction algorithm and a signal processing method incorporating wavelet decomposition and deep convolutional neural networks, repetitive features are extracted from the original underwater manifold signal to accurately describe the manifold's operating conditions. Leakage features are extracted from the aliased signal generated under the influence of various factors such as wave current and noise, obtaining the true leakage signal waveform. A fully electrically controlled pressure wave generator is developed to achieve leakage detection and location in underwater production manifolds based on transient excitation signal enhancement. Attached Figure Description

[0024] Figure 1This is a top view of the structure of a fully electrically controlled pressure wave generator;

[0025] Figure 2 It is an axial cross-sectional view of a fully electrically controlled pressure wave generator and a schematic diagram of an underwater manifold leak location system based on a fully electrically controlled pressure wave generator;

[0026] In the diagram, 101 is the generator rear end cover, 102 is the bearing housing, 103 is the reducer, 104 is the servo motor, 105 is the sealing cylinder, 106 is the front flange, 107 is the plunger rod, 108 is the load connection thread, 109 is the plunger, 110 is the upper end cover of the bearing, 111 is the synchronous pulley, 112 is the bearing assembly, 113 is the coupling, 114 is the buffer pad, 115 is the conduit, 116 is the ball screw, 117 is the dynamic seal ring, 118 is the underwater high-pressure sealing chamber, 201 is the underwater production manifold signal acquisition and processing subsystem, 202 is the experimental pipe, and 203 is the upstream signal of the experimental pipe. Module 104: Leakage Detection Module for Experimental Pipe; Module 205: Downstream Signal Acquisition Module for Experimental Pipe; Module 206: Underwater Production Manifold Pipeline; Module 207: Upstream Signal Acquisition Module for Underwater Production Manifold Pipeline; Module 208: Leakage Detection Module for Underwater Production Manifold Pipeline; Module 209: Downstream Signal Acquisition Module for Underwater Production Manifold Pipeline; Module 210: Underwater Production Manifold Signal Processing Unit; Module 211: Signal Feature Extraction Module; Module 212: Signal Denoising Module; Module 213: Leakage Location Subsystem for Underwater Production Manifold; Module 214: Time Difference Calculation Module; Module 215: Signal Reconstruction Module; Module 216: Leakage Location Module. Detailed Implementation

[0027] like Figure 1 The image shown is a top view of the fully electrically controlled pressure wave generator; as shown... Figure 2 As shown, an underwater manifold leakage location method and system based on a fully electrically controlled pressure wave generator includes three steps: I. Establishment of the fully electrically controlled pressure wave generator structure subsystem; II. Establishment of the underwater production manifold signal acquisition and processing subsystem; III. Establishment of the underwater production manifold leakage location subsystem.

[0028] The specific steps for establishing the fully electrically controlled pressure wave generator subsystem are as follows:

[0029] A101: Establish a fully electrically controlled pressure wave generator transmission system. Based on the geometry and working pressure of the underwater production manifold, select the drive and speed control devices for the fully electrically controlled pressure wave generator, and design the transmission structure to convert the motor's rotary motion into the lead screw's rotary motion;

[0030] A102: Set the pressure waveform generated by the fully electrically controlled pressure wave generator. For the pressure in the underwater production manifold, the plunger speed is set through expert decision-making and simulation experiments. The motor drives the plunger to achieve reciprocating linear motion through the transmission mechanism, generating a pressure wave. This wave is installed on the manifold through a flange structure before the return, impacting the fluid in the underwater production manifold and generating a transient excitation signal to amplify the leakage signal of the underwater production manifold.

[0031] A103: High-pressure sealing of the fully electrically controlled pressure wave generator. The fully electrically controlled pressure wave generator is installed entirely in the underwater high-pressure sealed chamber 118, and the drive module is installed in the underwater control module, controlled by the main control station on the platform via an umbilical cable.

[0032] The specific steps for establishing the underwater production manifold signal acquisition and processing subsystem are as follows:

[0033] A201: Establishment of the underwater production manifold signal acquisition system. A test pipe is installed on the underwater production manifold. The test pipe is internally polished, tightly fitted with the plunger to prevent leakage, and its geometric parameters are consistent with those of the underwater production manifold. An upstream signal acquisition module 203, a leakage detection module 204, and a downstream signal acquisition module 205 are installed on the test pipe 202. An upstream signal acquisition module 207, a leakage detection module 208, and a downstream signal acquisition module 209 are installed on the underwater production manifold pipeline 206 to acquire various signals from the underwater production manifold.

[0034] A202: Calculation of pressure wave propagation velocity in underwater production manifolds.

[0035] ① Using the upstream signal acquisition module 203, the downstream signal acquisition module 205, and the leakage generation module 204 of the experimental tube, a leakage was simulated. The transient change curve of the pipeline pressure wave at the known leakage point was analyzed. The arrival time of the known leakage point and the distance between the two signal acquisition modules of the experimental tube were extracted. The underwater production manifold pressure wave propagation speed Vc under the impact of the fully electronically controlled pressure wave generator was obtained through experiments.

[0036] ② The theoretical calculation formula is introduced to correct the calculation results of the pressure wave propagation velocity on the experimental tube. The average inner diameter of the tube is D, the wall thickness is δ, the flow area is ω, the liquid density is ρ, the liquid velocity is V, the bulk modulus of the transported fluid is K, the elastic modulus of the tube material is E, and the Poisson factor of the tube is μ. The time Δt required for the pressure wave peak to travel ΔL in the tube is analyzed to determine the pressure wave propagation velocity Va in the tube.

[0037] Based on the law of conservation of mass, the continuity of liquid, and the compression of liquid and expansion of pipe wall, the mass of liquid flowing into and out of the infinitesimal pipe segment within a time interval Δt is calculated. The mass of liquid flowing into the infinitesimal pipe segment within the time interval Δt can be expressed as:

[0038] m1=ρωVΔt

[0039] The mass of the outflowing micro-segment of liquid can be expressed as:

[0040] m2=(ρ+Δρ)(ω+Δω)(V-ΔV)Δt

[0041] Ignoring higher-order infinitesimals in the equation, the mass difference of the liquid entering and exiting the infinitesimal pipe segment within time Δt is:

[0042] Δm=m1-m2=ρωΔVΔt

[0043] The change in liquid mass within this infinitesimal pipe segment is examined based on liquid compression and pipe wall expansion. The mass before the change is:

[0044] m'1=ρωΔL

[0045] The mass after time Δt is:

[0046] m'2=(ρ+Δρ)(ω+Δω)ΔL

[0047] Ignoring higher-order infinitesimals in the equation, the mass difference of the liquid within the infinitesimal pipe segment over time Δt is obtained as follows:

[0048] Δm'=m'2-m'1=(ρΔω+ωΔρ)ΔL

[0049] An equation is established using the mass changes obtained from the two calculation methods. Replacing ΔL / Δt with Va, and simplifying, we get:

[0050]

[0051] Substituting the head conversion formula ΔH=VaΔV / g, we get:

[0052]

[0053] The calculation process for the circulation area change rate term and the density change rate term in the above formula is as follows:

[0054] Over a unit length, the circumferential tensile force T on the pipe wall under the action of the water head ΔH is (DρgΔH) / 2, where g is the gravity coefficient; the circumferential strain ε of the pipe wall... T Given T / (Eδ); the radial deformation is obtained as (Dε T Therefore, the area change rate term can be simplified to: ) / 2

[0055]

[0056] The rate of change of a liquid's density is related to its compressibility and the pressure applied to it. The compressibility of a liquid is measured using its bulk modulus K, expressed as:

[0057]

[0058] Therefore, we can conclude that:

[0059]

[0060] Substituting the above equation, we can obtain the formula for the propagation speed of pressure waves:

[0061]

[0062] Based on the actual physical constraints of the underwater production manifold, a pipeline correction coefficient is selected to establish a pressure wave propagation velocity correction model. In actual engineering, the underwater production manifold is installed in a supporting structure, so a pipeline correction coefficient ψ is introduced.

[0063]

[0064] The formula for calculating the propagation speed of pressure waves is:

[0065]

[0066] ③ Based on the underwater production manifold pressure wave propagation velocity Vc obtained from experiments on the experimental tube and the pressure wave propagation velocity Va calculated by the theoretical formula, the weighting coefficient p is determined through multiple sets of experiments and fitting. i q i The propagation velocity V of the underwater production manifold leakage pressure wave under the impact of the fully electrically controlled pressure wave generator was obtained. q :

[0067] V q =p i Vc+q i Va

[0068] A203: Establishment of the underwater production manifold signal processing unit. Signal processing mainly includes signal dimensionality reduction, feature extraction, signal denoising, and reconstruction of the acquired underwater production manifold pressure signals.

[0069] ① A signal dimensionality reduction algorithm based on improved principal component analysis is proposed. New variables are constructed using linear transformations of the extracted original feature parameters. Principal components are then extracted from these variables for signal dimensionality reduction. The original features are projected onto the dimension with the maximum projective information content, ensuring minimal information loss after dimensionality reduction. The extracted signal dataset is X. z :

[0070]

[0071] For X z Conduct standardized analysis, μ pi and σ pi Corresponding to x in sequence i Mean and standard deviation. Where x ij It is the specific value of the j-th feature corresponding to the i-th sample.

[0072]

[0073] Solve for X z Correlation coefficient matrix R:

[0074]

[0075] Find the eigenvalues ​​and eigenvectors of R:

[0076] |R-λI b |=0

[0077] Solving for b eigenvalues ​​yields λ1, λ2, ..., λb. b Solve the system of equations R×p=λ p This yields the eigenvectors P1, P2, ..., P corresponding to the eigenvalues. b The principal components are obtained by linearly combining the b eigenvectors of the data matrix Z, in the following form:

[0078]

[0079] F1 is the first principal component, F2 is the second principal component, and F... i i is the principal component.

[0080] Extract the optimal number of principal components and establish the cumulative contribution rate as the criterion, satisfying:

[0081]

[0082] ② A signal processing method based on the fusion of wavelet decomposition and deep convolutional neural network is proposed. A deep convolutional neural network model fused with wavelet decomposition is constructed.

[0083] The underwater production manifold pressure signal after principal component analysis is segmented according to a fixed sample length to obtain a sufficient number of one-dimensional sample sequences. These normalized sample sequences are then fed into a preprocessing layer for repeated preprocessing, converting them into two-dimensional vector inputs, with each vector having the same length as the initial vector. These two-dimensional vectors are then fed into a wavelet decomposition module to model each vector signal.

[0084] x(n) = s(n) + u(n)

[0085] Here, s(n) is the useful signal, and u(n) is the noise sequence. The submersible production manifold's leak-free pressure signal is pre-acquired and analyzed to distinguish noise components, forming a noise set during normal manifold operation. This set is then expanded and trained to obtain u(n).

[0086] After performing wavelet transform on the signal, we get:

[0087] WT x (a,b)=WT s (a,b)+WT u (a,b)

[0088] u(n) is a stationary random signal with zero mean and independent distribution, denoted as:

[0089] u=(u(0)u(1)…u(N-1)) T

[0090] Then we have:

[0091]

[0092] In the formula, E{·} represents the mean operation, and Q is the covariance matrix of u.

[0093] W is the wavelet transform matrix, x and s are the vectors corresponding to x(n) and s(n), and vectors X, S, and U are the wavelet transforms of x(n), s(n), and u(n), respectively. P is the covariance matrix of U, so we can obtain:

[0094] P=E{UU T}=E{Wuu T W T}=WQW T

[0095] The processed signal is analyzed to obtain low-frequency and high-frequency components. These two components are concatenated and fed into a convolutional module for automatic feature extraction. The convolutional module consists of convolutional layers, batch normalization (BN) layers, and ReLU activation layers. The convolutional layers extract important features in both the time and frequency domains, the BN layers improve network training efficiency, and the ReLU layers set near-zero features to 0, improving network sparsity and saving computational resources. The convolutional module learns useful pressure feature information from each frequency band component. The extracted feature components are then fed into a wavelet decomposition layer for wavelet decomposition, obtaining multi-frequency information for multiple feature components. These components are then combined and concatenated again before being fed back into the convolutional module. The wavelet decomposition layer and the convolutional layer execute alternately. After this loop is completed, the convolutional module is used for repeated feature extraction, achieving feature extraction, signal denoising, and signal reconstruction of the underwater production manifold leakage pressure signal.

[0096] The specific steps for establishing the underwater production manifold leak location subsystem are as follows:

[0097] A301: Calculation of the arrival time difference of transient pressure signal in the underwater production manifold. Based on the analysis of the transient pressure signal processed by A203, leakage inflection point information is identified, and the signal acquisition module closest to the leak point is automatically matched. The time it takes for the transient pressure drop to reach the pressure transmitter is extracted. The distance L between the upstream signal acquisition module 207 and the downstream signal acquisition module 209 of the underwater production manifold is measured. Based on the signal processed by A203, the times when the upstream and downstream sensors receive the pressure wave generated at the leak point are selected as t. q1 With t q2 Accurately extract the time difference t between the transient pressure changes acquired by the upstream signal acquisition module 207 and the downstream signal acquisition module 209 of the underwater production manifold. 12 ;

[0098] A302: Leak location in underwater production manifolds. Based on the pressure wave propagation velocity V of the leak in the underwater production manifold. q The distance L between the upstream signal acquisition module 207 and the downstream signal acquisition module 209 of the underwater production manifold and the transient pressure wave arrival time difference t. 12 Calculate the location of the leak point, with fluid velocity v. g We can obtain:

[0099]

[0100]

[0101] If the distance between pressure sensor A and the leak point is Xg, then the formula for locating the leak in the underwater production manifold based on transient pressure signal enhancement is:

[0102]

[0103] like Figure 2As shown, an underwater manifold leakage location method and system based on a fully electrically controlled pressure wave generator includes a fully electrically controlled pressure wave generator structural subsystem, an underwater production manifold signal acquisition and processing subsystem, and an underwater production manifold leakage location subsystem. The structure of the fully electrically controlled pressure wave generator includes a generator rear end cover 101, a bearing housing 102, a reducer 103, a servo motor 104, a sealing cylinder 105, a front flange 106, a plunger rod 107, a load-connecting thread 108, a plunger 109, a bearing upper end cover 110, a synchronous pulley 111, a bearing assembly 112, a coupling 113, a buffer pad 114, a conduit 115, a ball screw 116, a dynamic sealing ring 117, and an underwater high-pressure sealing chamber 118. The servo motor 104 is connected to the reducer 103, which provides power to the fully electronically controlled pressure wave generator and regulates its speed. The servo motor 104 and the reducer 103 are fixed by the bearing housing 102. The synchronous pulley 111 connects the servo motor 104 and the coupling 113, which is sealed by the generator's rear end cover 101. The coupling 113 is connected to the ball screw 116, which converts the motor's rotational motion into the screw's rotational motion. The bearing assembly 112 is axially fixed by the upper bearing cover 110, which restricts the axial movement of the ball screw 116 and ensures rotational accuracy. The ball screw 116 moves axially in the guide tube 115, driving the plunger rod 10. 7. The ball screw 116 performs reciprocating linear motion. The buffer pad 114 is used to absorb the impact and vibration during the motion, ensuring the accuracy and stability of the ball screw 116. The whole is sealed by the sealing cylinder 105. The end of the plunger rod 107 is connected to the plunger 109 with a load-connecting thread 108 to achieve reciprocating linear motion, generating transient excitation in the pipeline and emitting pressure waves to enhance the leakage signal. The dynamic sealing ring 117 is used for dynamic sealing during the reciprocating linear motion of the plunger rod 107. The entire structure of the pressure wave generator is placed in the underwater high-pressure sealing chamber 118. The front flange 106 is connected to the flange of the experimental tube 202 for the installation of the pressure wave generator.

[0104] The underwater production manifold signal acquisition and processing subsystem 201 includes an experimental pipe 202, an underwater production manifold pipeline 206, and an underwater production manifold signal processing unit 210; the experimental pipe 202 includes an upstream signal acquisition module 203, an experimental pipe leakage detection module 204, and an downstream signal acquisition module 205; the underwater production manifold pipeline 206 includes an upstream signal acquisition module 207, an underwater production manifold pipeline leakage detection module 208, and an underwater production manifold pipeline downstream signal acquisition module 209; the underwater production manifold signal processing unit 210 includes a signal feature extraction module 211, a signal denoising module 212, and a signal reconstruction module 215. The front flange 106 is connected to the flange of the experimental tube 202. The interior of the experimental tube 202 is polished and fits tightly with the plunger to prevent leakage. An upstream signal acquisition module 203, a leakage detection module 204, and a downstream signal acquisition module 205 are installed on the experimental tube 202 to simulate leakage and acquire pipeline signals. The experimental tube 202 is connected to the flange of the underwater production manifold 206. The upstream signal acquisition module 207 and the downstream signal acquisition module 209 of the underwater production manifold acquire leakage signals from the leakage detection module 208. The signals acquired by block 203, downstream signal acquisition module 205 of the experimental pipe, upstream signal acquisition module 207 of the underwater production manifold, and downstream signal acquisition module 209 of the underwater production manifold are transmitted to the underwater production manifold signal processing unit 210 via power lines. The underwater production manifold signal processing unit 210 includes a signal feature extraction module 211, a signal denoising module 212, and a signal reconstruction module 215, which are used for feature extraction, denoising, and reconstruction of the original signals extracted by the dynamic pressure transmitter. The signal processing results of the underwater production manifold signal processing unit 210 are transmitted to the underwater production manifold leak location subsystem 213.

[0105] The underwater production manifold leakage location subsystem 213 includes a time difference calculation module 214 and a leakage location module 216. The time difference calculation module 214 is used to calculate the time difference of arrival of the transient pressure wave of the leakage signal. The leakage location module 216 is connected to the time difference calculation module 214 through a cable, and takes in the arrival time difference and the propagation speed of the pressure wave to achieve accurate location of the leakage point of the underwater production manifold.

Claims

1. A method for locating leaks in underwater manifolds based on a fully electrically controlled pressure wave generator, characterized in that: It includes three steps: the establishment of the fully electronically controlled pressure wave generator structure subsystem, the establishment of the underwater production manifold signal acquisition and processing subsystem, and the establishment of the underwater production manifold leakage location subsystem; The specific steps for establishing the fully electrically controlled pressure wave generator subsystem are as follows: A101: Establish a fully electrically controlled pressure wave generator transmission system. Based on the geometry and working pressure of the underwater production manifold, select the drive device and speed control device of the fully electrically controlled pressure wave generator, and design the transmission structure to convert the rotational motion of the motor into the rotational motion of the lead screw. A102: Set the pressure waveform generated by the fully electronically controlled pressure wave generator. Based on the pressure of the underwater production manifold, the plunger speed is set through expert decision-making and simulation experiments. The motor drives the plunger to achieve reciprocating linear motion through the transmission mechanism, generating a pressure wave. It is installed on the manifold through the flange structure before the return, impacting the fluid in the underwater production manifold and generating a transient excitation signal to enhance the leakage signal of the underwater production manifold. A103: The fully electrically controlled pressure wave generator is sealed under high pressure. The entire fully electrically controlled pressure wave generator is installed in an underwater high-pressure sealed chamber, and the drive module is installed in an underwater control module. It is controlled by the main control station on the platform via an umbilical cable. The specific steps for establishing the underwater production manifold signal acquisition and processing subsystem are as follows: A201: Establishment of an underwater production manifold signal acquisition system; a test pipe is installed on the underwater production manifold, with its interior polished to ensure a tight fit with the plunger and prevent leakage, and all geometric parameters are consistent with those of the underwater production manifold; an upstream signal acquisition module, a leak detection module, and a downstream signal acquisition module are installed on the test pipe; an upstream signal acquisition module, a leak detection module, and a downstream signal acquisition module are installed on the underwater production manifold pipeline to acquire various signals from the underwater production manifold; A202: Calculation of pressure wave propagation velocity in underwater production manifolds; Based on the pressure wave propagation velocity Vc obtained from experiments on the experimental tube and the pressure wave propagation velocity Va calculated by the theoretical formula, the weighting coefficient p was determined through fitting multiple sets of experiments and based on the experimental results. i q i The propagation velocity V of the underwater production manifold leakage pressure wave under the impact of the fully electrically controlled pressure wave generator was obtained. q : A203: Establishment of the underwater production manifold signal processing unit; The specific steps for establishing the underwater production manifold leak location subsystem are as follows: A301: Calculation of the arrival time difference of transient pressure signals in underwater production manifolds; analysis of the transient pressure signals processed by A203 to identify leakage inflection point information, automatic matching of the signal acquisition module closest to the leakage point, and extraction of the time it takes for the transient pressure drop to reach the pressure transmitter; measurement of the distance L between the upstream and downstream signal acquisition modules of the underwater production manifold, and selection of the time t for the upstream and downstream sensors to receive the pressure wave generated at the leakage point based on the signals processed by A203. q1 With t q2 Accurately extract the time difference t between the transient pressure changes acquired by the upstream signal acquisition module and the downstream signal acquisition module of the underwater production manifold. 12 ; A302: Leak location in underwater production manifolds, based on the propagation velocity V of the leaking pressure wave in the underwater production manifold. q The distance L between the upstream signal acquisition module and the downstream signal acquisition module of the underwater production manifold, and the transient pressure wave arrival time difference t. 12 Calculate the location of the leak point, with fluid velocity v. g We can obtain: If the distance between pressure sensor A and the leak point is Xg, then the formula for locating the leak in the underwater production manifold based on transient pressure signal enhancement is: The underwater manifold leakage location method based on the fully electrically controlled pressure wave generator is applied to the underwater manifold leakage location system based on the fully electrically controlled pressure wave generator. The system consists of three parts: the fully electrically controlled pressure wave generator structure subsystem, the underwater production manifold signal acquisition and processing subsystem, and the underwater production manifold leakage location subsystem. The structural subsystem of the fully electrically controlled pressure wave generator includes the generator rear end cover, bearing housing, reducer, servo motor, sealing cylinder, front flange, plunger rod, load connection thread, plunger, bearing upper end cover, synchronous pulley, bearing assembly, coupling, buffer pad, guide tube, ball screw, dynamic seal ring and underwater high-pressure sealing chamber. The underwater production manifold signal acquisition and processing subsystem includes an experimental pipe, an underwater production manifold pipeline, and an underwater production manifold signal processing unit. The experimental pipe includes an upstream signal acquisition module, a leak detection module, and a downstream signal acquisition module. The underwater production manifold pipeline includes an upstream signal acquisition module, a leak detection module, and a downstream signal acquisition module. The underwater production manifold signal processing unit includes a signal feature extraction module, a signal denoising module, and a signal reconstruction module. The underwater production manifold leak location subsystem includes a time difference calculation module and a leak location module.

2. The underwater manifold leakage location method based on a fully electrically controlled pressure wave generator according to claim 1, characterized in that: Calculation of pressure wave propagation velocity in underwater production manifolds: ① Using the upstream signal acquisition module, the downstream signal acquisition module, and the leakage generation module of the experimental tube to simulate leakage, analyze the transient change curve of the pipeline pressure wave at the known leakage point, extract the arrival time of the known leakage point and the distance between the two signal acquisition modules of the experimental tube, and calculate the underwater production manifold pressure wave propagation speed Vc under the impact of the fully electronically controlled pressure wave generator through experiments. ② The theoretical calculation formula is introduced to correct the calculation results of the pressure wave propagation velocity on the experimental tube. The average inner diameter of the tube is D, the wall thickness is δ, the flow area is ω, the liquid density is ρ, the liquid velocity is V, the bulk modulus of the transported fluid is K, the elastic modulus of the tube material is E, and the Poisson factor of the tube is μ. The time Δt required for the pressure wave peak to travel ΔL in the tube is analyzed to determine the pressure wave propagation velocity Va in the tube. Based on the law of conservation of mass, the continuity of liquid, and the compression of liquid and expansion of pipe wall, the mass of liquid flowing into and out of the infinitesimal pipe segment within a time interval Δt is calculated. The mass of liquid flowing into the infinitesimal pipe segment within the time interval Δt can be expressed as: The mass of the outflowing micro-segment of liquid can be expressed as: Ignoring higher-order infinitesimals in the equation, the mass difference of the liquid entering and exiting the infinitesimal pipe segment within time Δt is: The change in liquid mass within this infinitesimal pipe segment is examined based on liquid compression and pipe wall expansion. The mass before the change is: The mass after time Δt is: Ignoring higher-order infinitesimals in the equation, the mass difference of the liquid within the infinitesimal pipe segment over time Δt is obtained as follows: An equation is established using the mass changes obtained from the two calculation methods. Replacing ΔL / Δt with Va, and simplifying, we get: Substituting the head conversion formula ΔH=VaΔV / g, we get: The calculation process for the circulation area change rate term and the density change rate term in the above formula is as follows: Over a unit length, the circumferential tensile force T on the pipe wall under the action of the water head ΔH is (DρgΔH) / 2, where g is the gravity coefficient; the circumferential strain ε of the pipe wall... T Given T / (Eδ); the radial deformation is obtained as (Dε T Therefore, the area change rate term can be simplified to: )⁄2 The rate of change of a liquid's density is related to its compressibility and the pressure applied to it. The compressibility of a liquid is measured using its bulk modulus K, expressed as: Therefore, we can conclude that: Substituting the above equation, we can obtain the formula for the propagation speed of pressure waves: Based on the actual physical constraints of the underwater production manifold, a pipeline correction coefficient is selected to establish a pressure wave propagation velocity correction model. In actual engineering, the underwater production manifold is installed in a supporting structure, so a pipeline correction coefficient ψ is introduced. The formula for calculating the propagation speed of pressure waves is: ③ Based on the underwater production manifold pressure wave propagation velocity Vc obtained from experiments on the experimental tube and the pressure wave propagation velocity Va calculated by the theoretical formula, the weighting coefficient p is determined by fitting multiple sets of experiments and the results. i q i The propagation velocity V of the underwater production manifold leakage pressure wave under the impact of the fully electrically controlled pressure wave generator was obtained. q : 。 3. The underwater manifold leakage location method based on a fully electrically controlled pressure wave generator according to claim 1, characterized in that: The underwater production manifold signal processing unit was established. Signal processing mainly includes signal dimensionality reduction, feature extraction, signal denoising, and reconstruction of the acquired raw pressure signals from the underwater production manifold. ① A signal dimensionality reduction algorithm based on improved principal component analysis is proposed. This algorithm constructs new variables using linear transformations of the extracted original feature parameters, extracts principal components from these variables, and performs signal dimensionality reduction. The original features are projected onto the dimension with the maximum projective information content, ensuring minimal information loss after dimensionality reduction. The extracted signal dataset is X. z : For X z Conduct standardized analysis, μ pi and σ pi Corresponding to x in sequence i Mean and standard deviation, where x ij It is the specific value of the j-th feature corresponding to the i-th sample; Solve for X z Correlation coefficient matrix R: Find the eigenvalues ​​and eigenvectors of R: Solving for b eigenvalues ​​yields λ1, λ2, ..., λb. b Solve the system of equations R×p=λ p This yields the eigenvectors P1, P2, ..., P corresponding to the eigenvalues. b The principal components are obtained by linearly combining the b eigenvectors of the data matrix Z, in the following form: F1 is the first principal component, F2 is the second principal component, and F... i It is the i-th principal component; Extract the optimal number of principal components and establish the cumulative contribution rate as the criterion, satisfying: ② A signal processing method based on the fusion of wavelet decomposition and deep convolutional neural network is proposed, and a deep convolutional neural network model fused with wavelet decomposition is constructed. The underwater production manifold pressure signal after principal component analysis is segmented according to a fixed sample length to obtain a sufficient number of one-dimensional sample sequences. The normalized sample sequences are then fed into a preprocessing layer for repeated preprocessing, converting them into two-dimensional vector inputs. The length of each vector is consistent with the initial length. The two-dimensional vectors are then fed into a wavelet decomposition module to model each vector signal. Where s(n) is the useful signal and u(n) is the noise sequence. The underwater production manifold leak-free pressure signal is collected in advance for analysis, the noise components are distinguished, the noise set when the manifold is working normally is formed and expanded for training to obtain u(n). After performing wavelet transform on the signal, we obtain: u(n) is a stationary random signal with zero mean and independent distribution, denoted as: Then we have: In the formula, E{·} represents the mean operation, and Q is the covariance matrix of u; W is the wavelet transform matrix, x and s are the vectors corresponding to x(n) and s(n), and vectors X, S, and U are the wavelet transforms of x(n), s(n), and u(n), respectively. P is the covariance matrix of U, so we can obtain: The processed signal is analyzed to obtain low-frequency and high-frequency components. These two components are concatenated and fed into a convolutional module for automatic feature extraction. The convolutional module consists of convolutional layers, batch normalization (BN) layers, and ReLU activation layers. The convolutional layers extract important features in the time and frequency domains, the BN layers improve network training efficiency, and the ReLU layers set near-zero features to 0, improving network sparsity and saving computational resources. The convolutional module learns useful pressure feature information from each frequency band component. The extracted feature components are then fed into a wavelet decomposition layer for wavelet decomposition, obtaining multi-frequency information of multiple feature components. These components are then combined and concatenated again before being fed into the convolutional module. The wavelet decomposition layer and the convolutional layer are executed alternately. After this loop is completed, the convolutional module is used to perform repeated feature extraction, realizing feature extraction, signal denoising, and signal reconstruction of the underwater production manifold leakage pressure signal.

4. The underwater manifold leakage location method based on a fully electrically controlled pressure wave generator according to claim 1, characterized in that: A servo motor connects to a reducer, providing power and speed regulation for the fully electronically controlled pressure wave generator. The servo motor and reducer are fixed by bearing housings. A synchronous pulley connects the servo motor and a coupling, sealed by the generator's rear end cover. The coupling connects to a ball screw, converting the motor's rotational motion into the screw's rotational motion. The bearing assembly is axially fixed by the upper end cover, limiting the ball screw's axial movement and ensuring rotational accuracy. The ball screw moves axially within the guide tube, driving the plunger rod in reciprocating linear motion. A buffer pad absorbs impacts and vibrations during the motion, ensuring the ball screw's accuracy and stability. The entire assembly is sealed by a sealed cylinder. The plunger rod's end has a load-bearing thread, connecting to the plunger to achieve reciprocating linear motion, generating transient excitation within the tube and emitting pressure waves. A dynamic sealing ring provides dynamic sealing during the plunger rod's reciprocating linear motion. The entire pressure wave generator structure is housed in an underwater high-pressure sealed chamber, with the front flange connected to the experimental tube flange for installation.

5. The underwater manifold leakage location method based on a fully electrically controlled pressure wave generator according to claim 1, characterized in that: The front flange connects to the test pipe flange. The interior of the test pipe is polished and fits tightly with the plunger to prevent leakage. An upstream signal acquisition module, a leak detection module, and a downstream signal acquisition module are installed on the test pipe to simulate leaks and acquire pipeline signals. The test pipe connects to the underwater production manifold flange. The upstream and downstream signal acquisition modules acquire leakage signals from the leak detection module. The signals acquired by these modules are transmitted via power lines to the underwater production manifold signal processing unit. The signal processing unit includes a signal feature extraction module, a signal denoising module, and a signal reconstruction module for extracting, denoising, and reconstructing the features of the acquired raw signals. The signal processing results from the signal processing unit are transmitted via cable to the underwater production manifold leak location subsystem.

6. The underwater manifold leakage location method based on a fully electrically controlled pressure wave generator according to claim 1, characterized in that: The time difference calculation module is used to calculate the time difference of arrival of the transient pressure wave of the leakage signal; the leakage location module is connected to the time difference calculation module through a cable to locate the leakage point.

Images