Submarine pipeline health monitoring method, device and equipment based on multi-data fusion

Abstract

The invention provides a submarine pipeline health monitoring method, device and equipment based on multi-data fusion, and belongs to the technical field of pipeline monitoring. The method comprises the steps that a physical feature sequence and a light intensity space-time image collected by a multi-source sensor system in the application process of the submarine pipeline are obtained, and the physical feature sequence is used for representing overall vibration information of the submarine pipeline and strain information at multiple different positions; the light intensity space-time image is used for representing intensity change information of optical fibers arranged on the outer side wall of the submarine pipeline; and performing fusion prediction analysis on the physical feature sequence and the light intensity space-time image by adopting a health monitoring model introducing a dynamic channel attention mechanism to obtain a risk fault prediction result of the submarine pipeline. According to the invention, fusion prediction analysis can be carried out on the multi-modal data generated when the submarine pipeline is applied through a deep learning technology, so that health monitoring of the submarine pipeline is realized more comprehensively and accurately.

Description

Technical Field

[0001] The present application relates to the technical field of pipeline monitoring, and in particular to a method, device and equipment for submarine pipeline health monitoring based on multi-data fusion. Background Art

[0002] Submarine pipelines are sealed conduits used to continuously transport large quantities of oil and gas under the seabed. During operation, they must be monitored for defects caused by the complex marine environment to ensure safe transportation. However, traditional submarine pipeline monitoring methods rely solely on single-modal data analysis, resulting in inaccurate and unreliable health monitoring results. Summary of the Invention

[0003] The main purpose of this application is to propose a submarine pipeline health monitoring method, device and equipment based on multi-data fusion, which can use deep learning technology to perform fusion prediction and analysis on the multimodal data generated during the application of submarine pipelines, thereby realizing more comprehensive and accurate health monitoring of submarine pipelines.

[0004] To achieve the above objectives, one aspect of this application proposes a submarine pipeline health monitoring method based on multi-data fusion, comprising:

[0005] Acquiring a physical feature sequence and a light intensity spatiotemporal image collected by a multi-source sensor system during application of a submarine pipeline, wherein the physical feature sequence is used to characterize overall vibration information of the submarine pipeline and strain information at multiple different locations, and the light intensity spatiotemporal image is used to characterize intensity variation information of an optical fiber arranged on the outer wall of the submarine pipeline;

[0006] A health monitoring model that introduces a dynamic channel attention mechanism is used to perform fusion prediction analysis on the physical feature sequence and the light intensity spatiotemporal image to obtain a risk failure prediction result of the submarine pipeline.

[0007] Furthermore, the health monitoring model includes a first network, a second network, a third network, and a fourth network; the health monitoring model using the dynamic channel attention mechanism performs a fusion prediction analysis on the physical feature sequence and the light intensity spatiotemporal image, and obtains the risk failure prediction result of the submarine pipeline including:

[0008] Inputting the physical feature sequence into the first network to perform feature extraction to obtain first feature data;

[0009] Inputting the first feature data into the second network to perform dynamic channel attention enhancement to obtain second feature data;

[0010] Inputting the light intensity spatiotemporal image into the third network for feature extraction to obtain third feature data;

[0011] The second characteristic data and the third characteristic data are input into the fourth network for fusion prediction analysis to obtain a risk failure prediction result of the submarine pipeline.

[0012] Furthermore, the first network is formed based on a one-dimensional convolutional neural network architecture; inputting the physical feature sequence into the first network for feature extraction to obtain first feature data includes:

[0013] The one-dimensional convolutional neural network architecture is used to perform multiple convolution pooling operations on the physical feature sequence to obtain the first feature data.

[0014] Furthermore, the second network includes a first fully connected layer, a first activation layer, a second fully connected layer, a second activation layer, and a feature enhancement layer connected in sequence; the first feature data is input into the second network for dynamic channel attention enhancement to obtain second feature data including:

[0015] Using the first fully connected layer to perform channel compression on the first feature data, and then using the first activation layer to perform nonlinear activation on the channel-compressed first feature data to obtain initial feature data;

[0016] Performing channel restoration on the initial feature data using the second fully connected layer, and then normalizing the initial feature data after channel restoration using the second activation layer to obtain attention weights;

[0017] The feature enhancement layer is used to perform weighted summation on the attention weight and the first feature data to obtain the second feature data.

[0018] Furthermore, the third network is formed based on the ResNet50 network architecture; the inputting the light intensity spatiotemporal image into the third network for feature extraction to obtain third feature data includes:

[0019] The ResNet50 network architecture is used to perform multiple convolution and residual processing on the light intensity spatiotemporal image to obtain the third feature data.

[0020] Furthermore, the fourth network includes a splicing layer, a third fully connected layer, and a fourth fully connected layer connected in sequence; inputting the second feature data and the third feature data into the fourth network for fusion prediction analysis to obtain a risk failure prediction result of the submarine pipeline includes:

[0021] flattening and horizontally splicing the second feature data and the third feature data using the splicing layer to obtain fused feature data;

[0022] The third fully connected layer is used to integrate the fused feature data, and the fourth fully connected layer is used to regularize and classify the integrated fused feature data to obtain a risk failure prediction result of the submarine pipeline.

[0023] Furthermore, the multi-source sensor system includes an optical time domain reflectometer, a fiber Bragg grating vibration sensor and a plurality of fiber Bragg grating strain sensors; the light intensity spatiotemporal image is obtained based on the optical time domain reflectometer; and the physical feature sequence is obtained based on the fiber Bragg grating vibration sensor and the plurality of fiber Bragg grating strain sensors.

[0024] To achieve the above objectives, another aspect of the present application provides a submarine pipeline health monitoring device based on multi-data fusion, comprising:

[0025] An acquisition module is configured to acquire a physical feature sequence and a light intensity spatiotemporal image collected by a multi-source sensor system during application of a submarine pipeline. The physical feature sequence is used to characterize the overall vibration information of the submarine pipeline and strain information at multiple different positions. The light intensity spatiotemporal image is used to characterize intensity variation information of an optical fiber arranged on the outer wall of the submarine pipeline.

[0026] The processing module is used to use a health monitoring model that introduces a dynamic channel attention mechanism to perform fusion prediction analysis on the physical feature sequence and the light intensity spatiotemporal image to obtain a risk failure prediction result of the submarine pipeline.

[0027] To achieve the above-mentioned purpose, another aspect of the present application provides an electronic device, which includes a memory and a processor, wherein the memory stores a computer program, and the processor implements the above-mentioned method when executing the computer program.

[0028] To achieve the above-mentioned purpose, another aspect of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and the computer program implements the above-mentioned method when executed by a processor.

[0029] This application has at least the following beneficial effects: By acquiring multimodal data representing different feature types generated during submarine pipeline operation and then employing deep learning techniques for fusion and predictive analysis, this application leverages the rich information on pipeline variation characteristics to enable more comprehensive and accurate submarine pipeline health monitoring. Multimodal data fusion and intelligent analysis can better address the complex and changing submarine pipeline environment and enhance anti-interference capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 This is a flow chart of a submarine pipeline health monitoring method based on multi-data fusion provided in an embodiment of the present application;

[0031] Figure 2 This is a schematic diagram of the composition of the health monitoring model provided in the embodiment of the present application;

[0032] Figure 3 This is a schematic diagram of the structure of a submarine pipeline health monitoring device based on multi-data fusion provided in an embodiment of the present application;

[0033] Figure 4 This is a schematic diagram of the hardware structure of the electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0034] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application is further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain the present application and are not intended to limit the present application. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the embodiments of the present application. They are merely examples of devices and methods consistent with some aspects of the embodiments of the present application as detailed in the appended claims.

[0035] It will be understood that the terms "first", "second", etc. used in this application may be used herein to describe various concepts, but unless otherwise specified, these concepts are not limited by these terms. These terms are only used to distinguish one concept from another. For example, without departing from the scope of the embodiments of the present application, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the words "if" and "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0036] The terms "at least one", "plurality", "each", "any", etc. used in this application include "at least one", "two" or more, "plurality" or "each", "any" or "any one", "each" or "any one" as used herein.

[0037] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this application pertains. The terms used herein are for the purpose of describing the embodiments of this application only and are not intended to limit this application.

[0038] Submarine pipelines are sealed conduits used to continuously transport large quantities of oil and gas under the seabed. They are a key component of offshore oil and gas field development and production, and are the fastest, safest, most economical, and reliable means of transporting oil and gas at sea. Their advantages include continuous transportation, virtually unaffected by environmental conditions, and the elimination of production cuts or shutdowns due to capacity constraints at offshore storage facilities or delayed shuttle tanker pickups. Furthermore, submarine pipelines offer a short construction period, rapid commissioning, easy management, and low operating costs. Submarine pipelines have become a vital link and focal point for long-distance, cross-sea energy transmission and the large-scale utilization of renewable energy.

[0039] During the operation of submarine pipelines, it is necessary to monitor them for defects caused by the complex marine environment to ensure their safe transportation. However, traditional submarine pipeline monitoring methods rely solely on single-modal data analysis and judgment, making the resulting health monitoring results inaccurate and unreliable.

[0040] In light of this, embodiments of the present application provide a method, apparatus, and device for submarine pipeline health monitoring based on multi-data fusion. This solution acquires multimodal data representing different feature types generated during submarine pipeline operation and then employs deep learning techniques for fusion prediction and analysis. This solution fully utilizes the rich information on pipeline variation characteristics to achieve more comprehensive and accurate submarine pipeline health monitoring. Through multimodal data fusion and intelligent analysis, it can better cope with the complex and changing submarine pipeline environment and improve anti-interference capabilities.

[0041] The embodiment of the present application provides a submarine pipeline health monitoring method based on multi-data fusion, which relates to the field of pipeline monitoring technology and can be applied to a terminal, a server, or software running on a terminal or a server. In some embodiments, the terminal can be a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart watch, and a vehicle-mounted terminal, etc., but is not limited thereto; the server side can be configured as an independent physical server, or as a server cluster or distributed system composed of multiple physical servers, or as a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms. The server can also be a node server in a blockchain network; the software can be an application that implements the above method, etc., but is not limited to the above forms.

[0042] See also Figure 1 , Figure 1 This is an optional flow chart of a submarine pipeline health monitoring method based on multi-data fusion provided in an embodiment of the present application. The method may include, but is not limited to, the following steps S101 to S102:

[0043] Step S101: obtaining a physical feature sequence and a light intensity spatiotemporal image collected by a multi-source sensor system during application of a submarine pipeline;

[0044] Step S102: A health monitoring model that introduces a dynamic channel attention mechanism is used to perform fusion prediction analysis on the physical feature sequence and the light intensity spatiotemporal image to obtain a risk failure prediction result of the submarine pipeline.

[0045] In this application, deep learning technology is used to perform fusion prediction and analysis on the multimodal data generated when submarine pipelines are used, which can achieve more comprehensive and accurate health monitoring of submarine pipelines.

[0046] In step S101 of some embodiments, the multi-source sensor system includes an optical time domain reflectometer (OTDR), a fiber Bragg grating (FBG) vibration sensor, and multiple fiber Bragg grating strain sensors. The physical feature sequence is obtained based on the fiber Bragg grating vibration sensor and the multiple fiber Bragg grating strain sensors, and primarily characterizes the overall vibration information of the submarine pipeline and the strain information at multiple different locations. The light intensity spatiotemporal image is obtained based on the optical time domain reflectometer, and primarily characterizes the intensity variation information of the optical fiber arranged on the outer wall of the submarine pipeline. It should be noted that the number of multiple fiber Bragg grating strain sensors can be set according to the complexity of the environment in which the submarine pipeline is located, and this application does not impose any restrictions on this.

[0047] In addition, the multi-source sensor system also includes a data acquisition and processing device and a fiber Bragg grating demodulator. The data acquisition and processing device is electrically connected to the optical time domain reflectometer, and the fiber Bragg grating demodulator is electrically connected to the fiber Bragg grating vibration sensor and multiple fiber Bragg grating strain sensors.

[0048] Specifically, the optical time-domain reflectometer is fixedly installed on the outer wall of the submarine pipeline and mainly monitors the backward Rayleigh scattered light in the optical fiber. The internal refractive index of the optical fiber and the position of the scattering unit can reflect the condition and position of the optical fiber affected by external disturbances. The backward Rayleigh scattered light is guided into the photodetector through the same optical circulator. The photodetector converts the optical signal into an electrical signal. The data acquisition and processing equipment then collects the electrical signal at a sampling rate of 10kHz and performs phase demodulation to obtain a restored three-dimensional light intensity signal. Then, data in the time and space dimensions are selected to form the light intensity spatiotemporal image, which can provide richer feature information for subsequent detection of submarine pipeline anomalies.

[0049] Specifically, multiple fiber Bragg grating strain sensors are fixedly arranged at multiple different positions on the outer wall of the submarine pipeline, with a certain interval between each two adjacent fiber Bragg grating strain sensors. A fiber Bragg grating vibration sensor can be arranged on the outer wall of the submarine pipeline and at a position away from any fiber Bragg grating strain sensor to avoid mutual interference. Relevant vibration data of the submarine pipeline is collected by the fiber Bragg grating vibration sensor, and then demodulated and processed by a fiber Bragg grating demodulator at a sampling frequency of 100 Hz to obtain a vibration signal. Relevant strain data at multiple different positions of the submarine pipeline is collected by the multiple fiber Bragg grating strain sensors, and then demodulated and processed by a fiber Bragg grating demodulator at a sampling frequency of 100 Hz to obtain corresponding multiple strain signals. Then, the vibration signal and the multiple strain signals are merged to obtain the physical feature sequence.

[0050] In this application, considering that abnormal conditions such as strain, stress or vibration in submarine pipelines can be manifested as different characteristics on different types of sensors, by using optical time domain reflectometers, fiber Bragg grating strain sensors and fiber Bragg grating vibration sensors to collect data when applied to submarine pipelines, it is helpful to more accurately identify potential risks of submarine pipelines in the subsequent data analysis process.

[0051] In step S102 of some embodiments, see Figure 2 As shown, the health monitoring model includes a first network, a second network, a third network, and a fourth network. The input end of the first network serves as the first input end of the health monitoring model, the input end of the third network serves as the second input end of the health monitoring model, the output end of the first network is connected to the input end of the second network, the output end of the second network and the output end of the third network are both connected to the input end of the fourth network, and the output end of the fourth network serves as the output end of the health monitoring model. The internal implementation process of the health monitoring model may include, but is not limited to, the following steps S201 to S204:

[0052] Step S201: input the physical feature sequence into the first network for feature extraction processing to obtain first feature data;

[0053] Step S202: inputting the first feature data into the second network for dynamic channel attention enhancement processing to obtain second feature data;

[0054] Step S203: inputting the light intensity spatiotemporal image into the third network for feature extraction processing to obtain third feature data;

[0055] Step S204: input the second characteristic data and the third characteristic data into the fourth network for fusion prediction analysis processing to obtain a risk failure prediction result of the submarine pipeline.

[0056] In this application, by using different networks to extract features from different modal data and then performing fusion prediction analysis, the abnormal information of submarine pipelines can be effectively captured, and the comprehensiveness and reliability of submarine pipeline health monitoring can be improved; by introducing a dynamic channel attention mechanism to enhance the feature data extracted from the physical feature sequence, it helps to identify more obvious abnormal feature information.

[0057] In step S201 of some embodiments, the first network is formed based on a one-dimensional convolutional neural network (1DCNN) architecture, and the first feature data can be obtained by performing multiple convolution pooling operations on the physical feature sequence using the one-dimensional convolutional neural network architecture.

[0058] Optionally, the one-dimensional convolutional neural network architecture includes a first convolutional layer, a first pooling layer, a second convolutional layer, and a second pooling layer connected in sequence, and its internal implementation process includes the following:

[0059] The physical feature sequence includes a column of vibration feature data and five columns of strain feature data. First, the physical feature sequence is subjected to an initial convolution operation through the first convolution layer to obtain a first feature map with a size of 16*L. Secondly, the first feature map is subjected to an initial maximum pooling operation through the first pooling layer to obtain a second feature map with a size of 16*0.5L. Then, the second feature map is subjected to a second convolution operation through the second convolution layer to obtain a third feature map with a size of 32*0.5L. Finally, the third feature map is subjected to a second maximum pooling operation through the second pooling layer to obtain a fourth feature map with a size of 32*0.25L, and the fourth feature map is output as the first feature data. The size of the maximum pooling window used in the two successive pooling operations is 2 to reduce the length of the time dimension.

[0060] In this application, by utilizing the convolution kernel contained in the one-dimensional convolutional neural network to slide on the physical feature sequence, the local feature information containing abnormalities can be effectively captured.

[0061] In step S202 of some embodiments, see Figure 2 As shown, the second network includes a first fully connected layer, a first activation layer, a second fully connected layer, a second activation layer, and a feature enhancement layer connected in sequence. The input end of the first fully connected layer and the input end of the feature enhancement layer are both connected to the output end of the first network. Its internal implementation process includes the following:

[0062] First, the first feature data is channel compressed by the first fully connected layer, that is, the number of channels of the first feature data is compressed from T0 to T1 by setting the dimensionality reduction factor r, T1=T0 / r, T0=32, and r=4 is preferably set in this application; secondly, the first feature data after channel compression is nonlinearly activated by the first activation layer to obtain initial feature data, mainly using the ReLU (Rectified Linear Unit) activation function to achieve this; then the initial feature data is channel restored by the second fully connected layer, that is, the number of channels of the initial feature data is restored from T1 to T0; then the initial feature data after channel restoration is normalized by the second activation layer to obtain attention weight, mainly using the Sigmoid activation function to achieve this; finally, the attention weight is weighted and summed with the first feature data by the feature enhancement layer to obtain the second feature data.

[0063] In this application, by introducing a dynamic channel attention mechanism, the importance of feature data of different channels can be dynamically adjusted, thereby enhancing the expression ability of sensor feature data that fluctuates significantly in abnormal events.

[0064] In step S203 of some embodiments, the third network is formed based on the ResNet50 network architecture. The third feature data can be obtained by performing multiple convolution and residual processing on the light intensity spatiotemporal image using the ResNet50 network architecture.

[0065] In this application, the ResNet50 network can effectively fuse spatiotemporal information to better understand the dynamic changes in the image, and can extract features of different scales through a multi-level structure to capture detailed feature information from local to global.

[0066] In step S204 of some embodiments, see Figure 2 As shown, the fourth network includes a concatenation layer, a third fully connected layer, and a fourth fully connected layer connected in sequence. The input end of the concatenation layer is connected to the output end of the feature enhancement layer and the output end of the third network respectively. The internal implementation process includes the following:

[0067] First, the second feature data and the third feature data are flattened into one-dimensional tensors through the splicing layer and then horizontally spliced ​​to obtain fused feature data, which can ensure feature integrity while reducing the complexity of subsequent feature analysis. Then, the fused feature data is integrated through the third fully connected layer. Finally, the integrated fused feature data is regularized and classified and predicted through the fourth fully connected layer to obtain the risk failure prediction result of the submarine pipeline, which is used to reflect the displacement and / or swing of the submarine pipeline caused by one or more events including anchor hitting, anchor hook pulling, ocean current and tide.

[0068] In some embodiments, the pre-training process of the health monitoring model mentioned in step S102 is described as follows:

[0069] Acquire a target data set, the target data set including multiple target data groups corresponding to multiple different target submarine pipelines, each target data group corresponding to a target submarine pipeline including multiple target data subgroups, each target data subgroup including a target physical feature sequence and a target light intensity spatiotemporal image generated when the target submarine pipeline is applied, and the target data subgroup carries a corresponding risk fault category label;

[0070] The target data set is divided into a training set and a test set according to a preset ratio. The health monitoring model is forward-propagated and trained based on the training set. The health monitoring model is then back-propagated based on a preset loss function. At the same time, the Adam (Adaptive Moment Estimation) optimizer is used to update the weights and bias parameters in the network model. The preset loss function is preferably set to a cross-entropy loss function. The performance of the trained health monitoring model is then tested based on the test set.

[0071] See also Figure 3 , Figure 3 This is a schematic diagram of an optional structural composition of a submarine pipeline health monitoring device based on multi-data fusion provided in an embodiment of the present application, which can implement the above-mentioned submarine pipeline health monitoring method based on multi-data fusion. The device may include, but is not limited to, the following:

[0072] An acquisition module 301 is used to acquire a physical feature sequence and a light intensity spatiotemporal image collected by a multi-source sensor system during the application process of a submarine pipeline;

[0073] The physical feature sequence is used to characterize the overall vibration information of the submarine pipeline and the strain information at multiple different positions, and the light intensity spatiotemporal image is used to characterize the intensity variation information of the optical fiber arranged on the outer wall of the submarine pipeline;

[0074] The processing module 302 is configured to perform a fusion prediction analysis on the physical feature sequence and the light intensity spatiotemporal image using a health monitoring model that introduces a dynamic channel attention mechanism, so as to obtain a risk failure prediction result of the submarine pipeline.

[0075] It can be understood that the contents of the above method embodiments are all applicable to the embodiments of the present device, the functions specifically implemented by the embodiments of the present device are the same as the functions specifically implemented by the above method embodiments, and the beneficial effects achieved by the embodiments of the present device are also the same as the beneficial effects achieved by the above method embodiments.

[0076] An embodiment of the present application further provides an electronic device comprising a memory and a processor. The memory stores a computer program, and the processor, when executing the computer program, implements the aforementioned method for submarine pipeline health monitoring based on multi-data fusion. The electronic device may include any intelligent terminal, such as a tablet computer or an in-vehicle computer.

[0077] It can be understood that the contents of the above method embodiments are all applicable to the embodiments of the present device, the functions specifically implemented by the embodiments of the present device are the same as the functions specifically implemented by the above method embodiments, and the beneficial effects achieved by the embodiments of the present device are also the same as the beneficial effects achieved by the above method embodiments.

[0078] See also Figure 4 , Figure 4 The hardware structure of an electronic device according to another embodiment is shown. The electronic device includes:

[0079] The processor 401 may be implemented as a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is configured to execute relevant programs to implement the technical solutions provided in the embodiments of the present application.

[0080] The memory 402 can be implemented in the form of a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 402 can store an operating system and other application programs. When the technical solutions provided by the embodiments of the present application are implemented through software or firmware, the relevant program code is stored in the memory 402 and is called by the processor 401 to execute the technical solutions provided by the embodiments of the present application.

[0081] Input / output interface 403, used to implement information input and output;

[0082] Communication interface 404, used to implement communication interaction between this device and other devices, which can be achieved through wired means (such as USB, network cable, etc.) or wireless means (such as mobile network, WiFi, Bluetooth, etc.);

[0083] Bus 405 , which transmits information between various components of the device (e.g., processor 401 , memory 402 , input / output interface 403 , and communication interface 404 );

[0084] The processor 401 , the memory 402 , the input / output interface 403 and the communication interface 404 are connected to each other in communication within the device via a bus 405 .

[0085] An embodiment of the present application further provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the computer program implements the above-mentioned submarine pipeline health monitoring method based on multi-data fusion.

[0086] It can be understood that the contents of the above method embodiments are all applicable to the present storage medium embodiment, the functions specifically implemented by the present storage medium embodiment are the same as the functions specifically implemented by the above method embodiments, and the beneficial effects achieved by the present storage medium embodiment are also the same as the beneficial effects achieved by the above method embodiments.

[0087] The memory, as a non-transient computer-readable storage medium, can be used to store non-transient software programs and non-transient computer executable programs. In addition, the memory may include a high-speed random access memory and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory may optionally include a memory remotely arranged relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0088] The embodiments described in the embodiments of this application are intended to more clearly illustrate the technical solutions of the embodiments of this application and do not constitute a limitation on the technical solutions provided by the embodiments of this application. Those skilled in the art will appreciate that with the evolution of technology and the emergence of new application scenarios, the technical solutions provided in the embodiments of this application are also applicable to similar technical problems.

[0089] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of the present application, and may include more or fewer steps than shown in the figures, or a combination of certain steps, or different steps.

[0090] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, i.e., they may be located in one place or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of this embodiment.

[0091] Those skilled in the art will appreciate that all or some of the steps, devices, and functional modules / units in the methods disclosed above may be implemented as software, firmware, hardware, or appropriate combinations thereof.

[0092] The terms "first", "second", "third", "fourth", etc. (if any) in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequential order. It should be understood that the numbers used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, device, product or equipment that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or equipment.

[0093] It should be understood that in this application, "at least one (item)" means one or more, and "plurality" means two or more. "And / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. "At least one of the following items" or similar expressions refers to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.

[0094] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the above-mentioned units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0095] The units described above as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0096] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0097] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes multiple instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of various embodiments of the present application. The aforementioned storage medium includes various media that can store programs, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0098] The preferred embodiments of the present invention are described above with reference to the accompanying drawings, but are not intended to limit the scope of the present invention. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and essence of the present invention should be within the scope of the present invention.

Claims

1. A submarine pipeline health monitoring method based on multi-data fusion, characterized in that: include: Acquiring a physical feature sequence and a light intensity spatiotemporal image collected by a multi-source sensor system during application of a submarine pipeline, wherein the physical feature sequence is used to characterize overall vibration information of the submarine pipeline and strain information at multiple different locations, and the light intensity spatiotemporal image is used to characterize intensity variation information of an optical fiber arranged on the outer wall of the submarine pipeline; A health monitoring model that introduces a dynamic channel attention mechanism is used to perform fusion prediction analysis on the physical feature sequence and the light intensity spatiotemporal image to obtain a risk failure prediction result of the submarine pipeline.

2. The submarine pipeline health monitoring method based on multi-data fusion according to claim 1 is characterized in that: The health monitoring model includes a first network, a second network, a third network, and a fourth network. The health monitoring model using the dynamic channel attention mechanism performs a fusion prediction analysis on the physical feature sequence and the light intensity spatiotemporal image to obtain the risk failure prediction result of the submarine pipeline, which includes: Inputting the physical feature sequence into the first network to perform feature extraction to obtain first feature data; Inputting the first feature data into the second network to perform dynamic channel attention enhancement to obtain second feature data; Inputting the light intensity spatiotemporal image into the third network for feature extraction to obtain third feature data; The second characteristic data and the third characteristic data are input into the fourth network for fusion prediction analysis to obtain a risk failure prediction result of the submarine pipeline.

3. The submarine pipeline health monitoring method based on multi-data fusion according to claim 2 is characterized in that: The first network is formed based on a one-dimensional convolutional neural network architecture; inputting the physical feature sequence into the first network for feature extraction to obtain first feature data includes: The one-dimensional convolutional neural network architecture is used to perform multiple convolution pooling operations on the physical feature sequence to obtain the first feature data.

4. The submarine pipeline health monitoring method based on multi-data fusion according to claim 2 is characterized in that: The second network includes a first fully connected layer, a first activation layer, a second fully connected layer, a second activation layer, and a feature enhancement layer connected in sequence; the first feature data is input into the second network for dynamic channel attention enhancement to obtain second feature data including: Using the first fully connected layer to perform channel compression on the first feature data, and then using the first activation layer to perform nonlinear activation on the channel-compressed first feature data to obtain initial feature data; Performing channel restoration on the initial feature data using the second fully connected layer, and then normalizing the initial feature data after channel restoration using the second activation layer to obtain attention weights; The feature enhancement layer is used to perform weighted summation on the attention weight and the first feature data to obtain the second feature data.

5. The submarine pipeline health monitoring method based on multi-data fusion according to claim 2 is characterized in that: The third network is formed based on the ResNet50 network architecture; the light intensity spatiotemporal image is input into the third network for feature extraction to obtain third feature data, including: The ResNet50 network architecture is used to perform multiple convolution and residual processing on the light intensity spatiotemporal image to obtain the third feature data.

6. The submarine pipeline health monitoring method based on multi-data fusion according to claim 2 is characterized in that: The fourth network includes a splicing layer, a third fully connected layer, and a fourth fully connected layer connected in sequence; the second feature data and the third feature data are input into the fourth network for fusion prediction analysis to obtain the risk failure prediction result of the submarine pipeline, which includes: flattening and horizontally splicing the second feature data and the third feature data using the splicing layer to obtain fused feature data; The third fully connected layer is used to integrate the fused feature data, and the fourth fully connected layer is used to regularize and classify the integrated fused feature data to obtain a risk failure prediction result of the submarine pipeline.

7. The submarine pipeline health monitoring method based on multi-data fusion according to claim 1 is characterized in that: The multi-source sensor system includes an optical time domain reflectometer, a fiber Bragg grating vibration sensor and multiple fiber Bragg grating strain sensors. The light intensity spatiotemporal image is obtained based on the optical time domain reflectometer, and the physical feature sequence is obtained based on the fiber Bragg grating vibration sensor and the multiple fiber Bragg grating strain sensors.

8. A submarine pipeline health monitoring device based on multi-data fusion, characterized in that: include: An acquisition module is configured to acquire a physical feature sequence and a light intensity spatiotemporal image collected by a multi-source sensor system during application of a submarine pipeline. The physical feature sequence is used to characterize the overall vibration information of the submarine pipeline and strain information at multiple different positions. The light intensity spatiotemporal image is used to characterize intensity variation information of an optical fiber arranged on the outer wall of the submarine pipeline. The processing module is used to use a health monitoring model that introduces a dynamic channel attention mechanism to perform fusion prediction analysis on the physical feature sequence and the light intensity spatiotemporal image to obtain a risk failure prediction result of the submarine pipeline.

9. An electronic device, characterized in that: The electronic device includes a memory and a processor, the memory stores a computer program, and the processor implements the method according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images