Tension anomaly detection method and device for anchoring system on floating photovoltaic platform

Abstract

The invention provides a tension anomaly detection method and device for an anchoring system on a floating type photovoltaic platform, and the method comprises the steps: building an initial tension distribution topological graph based on the platform structure of the floating type platform and the arrangement information of each mooring point after collecting the tension data of each mooring point and the real-time wind wave direction data of the floating type platform in real time; mapping the tension data of each mooring point to the initial tension distribution topological graph; inputting the mapped tension distribution topological graph into a pre-trained anomaly recognition model, and recognizing abnormal nodes with abnormal tension distribution; and based on the mooring point position corresponding to the abnormal node and the real-time wind wave direction data, a failed target seabed anchor is positioned. According to the method, the early tension abnormity of the floating type photovoltaic platform anchoring system can be accurately identified and positioned in real time, and the safety and the operation and maintenance efficiency of the platform are remarkably improved.

Description

Technical Field

[0001] This application relates to the field of marine engineering technology, and more specifically, to a method and apparatus for detecting tension anomalies in an anchoring system on a floating photovoltaic platform. Background Technology

[0002] Floating photovoltaic platforms are typically anchored to the seabed using a mooring system, which generally consists of mooring points on the platform, mooring cables, and a seabed anchor. Environmental loads (such as wind, waves, and currents) are transmitted through the platform to the mooring cables, and ultimately resisted by the seabed anchor. To ensure platform stability, mooring points are often arranged symmetrically.

[0003] In existing technologies, the means of monitoring the safety status of mooring systems after installation are relatively limited. Conventional practices mainly rely on periodic manual inspections or simple tension threshold alarms. These methods have significant shortcomings: manual inspections cannot achieve real-time monitoring and have a delayed response; while simple threshold alarms can only be triggered when tension is severely exceeded, failing to identify abnormal tension patterns within the normal threshold range that occur due to uneven environmental load distribution, such as abnormal tension distribution at a particular anchor point due to potential anchoring failure. Therefore, existing technologies struggle to achieve early, accurate location and warning of latent faults in mooring systems. Summary of the Invention

[0004] The purpose of this application is to provide a method and apparatus for detecting tension anomalies in the mooring system of a floating photovoltaic platform. By constructing a tension distribution topology map and applying graph neural networks or residual analysis, it is possible to identify and locate early tension anomalies in the mooring system of a floating photovoltaic platform in real time and accurately, achieving a leap from passive alarm to active intelligent diagnosis, and significantly improving the platform's safety and operation and maintenance efficiency.

[0005] Firstly, a method for detecting tension anomalies in an anchoring system on a floating photovoltaic platform is provided. The anchoring system includes multiple mooring points symmetrically arranged on the floating platform and corresponding seabed anchors connected to each mooring point via mooring cables. The method may include: Real-time acquisition of tension data at each mooring point and real-time wind and wave direction data of the floating platform; Based on the platform structure and mooring point layout information of the floating platform, an initial tension distribution topology map is constructed, and the tension data of each mooring point is mapped to this initial tension distribution topology map; The mapped tension distribution topology is input into a pre-trained anomaly detection model to identify anomalous nodes with abnormal tension distributions. Based on the location of the mooring point corresponding to the abnormal node and the real-time wind and wave direction data, the failed target seabed anchor was located.

[0006] In one possible implementation, the tension data is measured by tension sensors installed at each mooring point; The real-time wind and wave direction data is measured by a wind vane and a wave vane installed on the floating platform.

[0007] In one possible implementation, based on the platform structure of the floating platform and the arrangement information of each mooring point, an initial tension distribution topology map is constructed, and the tension data of each mooring point is mapped to this initial tension distribution topology map, including: Obtain the position of each mooring point on the floating platform; Each mooring point is defined as a node, and based on the position of each mooring point on the floating platform, the nodes corresponding to spatially adjacent mooring points that have a direct mechanical coupling relationship through the platform structure are connected by edges to construct a static initial tension distribution topology. The tension data collected in real time at each mooring point is used as the node feature of the corresponding node and assigned a value to obtain a tension distribution topology map that reflects the tension transmission path of the mooring system in physical space.

[0008] In one possible implementation, the pre-trained anomaly detection model is constructed using a graph neural network; The mapped tension distribution topology is input into a pre-trained anomaly detection model to identify anomalous nodes with abnormal tension distributions, including: By aggregating the tension features of each node itself and its neighboring nodes through at least one layer of graph convolution operation in the graph neural network, the spatial dependencies between nodes in the tension distribution topology graph are learned. Based on the learned spatial dependencies, the confidence level of each node that there is an abnormal tension is output, and based on the confidence level of each node that there is an abnormal tension, the abnormal nodes are identified.

[0009] In one possible implementation, the pre-trained anomaly detection model is constructed through residual analysis; The mapped tension distribution topology is input into a pre-trained anomaly detection model to identify anomalous nodes with abnormal tension distributions, including: Based on the real-time wind and wave direction data, the expected tension data of each node in the tension distribution topology map under the current environmental load is predicted, and a baseline tension distribution topology map is generated. Calculate the residual between the real-time tension distribution topology map composed of real-time tension data and the reference tension distribution topology map; Nodes whose residual values ​​exceed a preset threshold are identified as abnormal nodes.

[0010] In one possible implementation, based on the mooring point location and real-time wind and wave direction data corresponding to the anomalous node, the failed target seabed anchor is located, including: Based on the real-time wind and wave direction data, the main force direction of the floating platform and the corresponding expected high-tension mooring point area are determined. The location of the abnormal node is compared with the expected high-tension mooring point area; If the abnormal node is located in the expected high-tension mooring point area and the tension data of the abnormal node is abnormally low, then the seabed anchor corresponding to the abnormal node is determined to be the failed target seabed anchor.

[0011] In one possible implementation, after real-time acquisition of tension data at each mooring point and real-time wind and wave direction data of the floating platform, the method further includes: All tension data and real-time wind and wave direction data collected simultaneously in each group are unified under the same timestamp.

[0012] Secondly, a tension anomaly detection device for an anchoring system on a floating photovoltaic platform is provided. The anchoring system includes multiple mooring points symmetrically arranged on the floating platform and corresponding seabed anchors connected to each mooring point via mooring cables. The device may include: The data acquisition unit is used to collect tension data at each mooring point and real-time wind and wave direction data of the floating platform. The building unit is used to construct an initial tension distribution topology map based on the platform structure and mooring point layout information of the floating platform; The mapping unit is used to map the tension data of each mooring point to the initial tension distribution topology map. The identification unit is used to input the mapped tension distribution topology map into the pre-trained anomaly identification model to identify abnormal nodes with abnormal tension distribution. The positioning unit is used to locate the failed target seabed anchor based on the location of the mooring point corresponding to the abnormal node and real-time wind and wave direction data.

[0013] Thirdly, an electronic device is provided, which includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, used to store computer programs; When a processor executes a program stored in memory, it implements any of the steps described in the first aspect above.

[0014] Fourthly, a computer-readable storage medium is provided, wherein a computer program is stored therein, and when executed by a processor, the computer program implements the steps of any of the methods described in the first aspect above.

[0015] This application provides a method and apparatus for detecting tension anomalies in a mooring system on a floating photovoltaic platform. The mooring system includes multiple mooring points symmetrically arranged on the floating platform and corresponding seabed anchors connected to each mooring point via mooring cables. The method collects tension data from each mooring point and real-time wind and wave direction data of the floating platform in real time. Based on the platform structure and the arrangement information of each mooring point, an initial tension distribution topology map is constructed, and the tension data of each mooring point is mapped to this initial tension distribution topology map. The mapped tension distribution topology map is input into a pre-trained anomaly recognition model to identify abnormal nodes with abnormal tension distribution. Based on the location of the mooring point corresponding to the abnormal node and the real-time wind and wave direction data, the failed target seabed anchor is located. This method can identify and locate early tension anomalies in the mooring system of a floating photovoltaic platform in real time and accurately, significantly improving the platform's safety and operational efficiency. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 A flowchart illustrating a method for detecting abnormal tension in an anchoring system on a floating photovoltaic platform, provided in an embodiment of this application; Figure 2 A schematic diagram of the structure of a tension anomaly detection device for an anchoring system on a floating photovoltaic platform provided in this application embodiment; Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0018] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application. Unless otherwise defined, the technical or scientific terms used in this application should have the ordinary meaning understood by those skilled in the art. The words "first," "second," and similar terms used in this application do not indicate any order, quantity, or importance, but are only used to distinguish different components. The words "comprising" or "including," etc., mean that the element or object preceding the word covers the element or object listed after the word and its equivalents, but do not exclude other elements or objects. The words "connected," "coupled," or "connected," etc., are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. "Up," "down," "left," "right," etc., are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0019] The tension anomaly detection method for the mooring system on a floating photovoltaic platform provided in this application can detect early latent tension anomalies at specific anchor points in the mooring system on a floating photovoltaic platform in real time and intelligently, and accurately locate potentially failed anchor points, thereby overcoming the shortcomings of existing monitoring methods that are lagging and inaccurate.

[0020] The tension anomaly detection method for the anchoring system on a floating photovoltaic platform provided in this application embodiment relies on the following hardware structure: (1) Floating photovoltaic platform: It is usually a large floating structure, such as a rigid or semi-rigid platform composed of multiple pontoons, used to support photovoltaic panels.

[0021] (2) The mooring system is the fixing device for the floating photovoltaic platform, which may include: mooring points, mooring cables, and seabed anchors. Mooring points are multiple connection points symmetrically arranged along the edge of the floating platform (e.g., the six corners of a hexagonal platform to ensure balanced force). Mooring cables are steel or synthetic fiber cables connecting each mooring point to the seabed anchor. The seabed anchor is fixed to the seabed, providing final tensile strength. The floating platform is connected to the mooring cables via the mooring points arranged on the platform; the mooring cables are connected to the seabed anchor, transmitting tension to the seabed; the seabed anchor interacts with the seabed, providing resistance.

[0022] (3) A sensor system, which may include tension sensors and environmental monitoring sensors. Wherein: Tension sensors: Installed at each mooring point or on the mooring cable, these sensors continuously and in real time measure the tension data of the corresponding mooring cable. The measurement data (usually electrical signals) is transmitted to the computing and processing unit via a data acquisition module.

[0023] Environmental monitoring sensors may include wind vanes and wave vanes installed on the platform to obtain real-time wind and wave conditions at the platform. The core data are wind direction and main wave direction (usually expressed as geographical azimuth, such as 0° representing due north).

[0024] Furthermore, it may also include optional auxiliary sensors, such as underwater acoustic detectors and platform tilt sensor arrays.

[0025] Underwater acoustic detectors (such as hydrophones) can be deployed on the seabed near key mooring points to listen for acoustic signals of specific frequencies generated at the connection between the mooring cable and the anchor chain due to abnormal friction, impact, etc.

[0026] The platform tilt sensor array can be distributed at key locations on the floating photovoltaic platform to measure the platform's roll and pitch angles with high precision.

[0027] (4) Calculation and processing unit: It can be an industrial-grade computer or remote server located on a floating photovoltaic platform, equipped with a processor and memory, for receiving all sensor data, executing the method of this application, and outputting diagnostic results and alarm information, etc.

[0028] The preferred embodiments of this application are described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit this application. Furthermore, the embodiments and features in the embodiments of this application can be combined with each other without conflict.

[0029] Figure 1 This is a flowchart illustrating a method for detecting abnormal tension in an anchoring system on a floating photovoltaic platform, as provided in an embodiment of this application. Figure 1 As shown, the method may include: Step S110: Collect tension data at each mooring point and real-time wind and wave direction data of the floating platform.

[0030] The calculation and processing unit receives real-time tension data from all tension sensors at a fixed sampling frequency (e.g., 1Hz) to form a real-time tension data sequence {T1,T2,...,Tn} (where n is the number of mooring points, T1 represents the tension of mooring point 1, and so on). At the same time, it reads the real-time wind and wave direction data θ_env from the anemometer and wave compass, which represents the direction of the environmental load that has the greatest impact on the platform, i.e., the direction of propagation of environmental wind and waves expressed in angular form.

[0031] Furthermore, to ensure the accuracy of subsequent analysis, the collected tension data can be preprocessed by filtering and denoising: a low-pass filter (such as a Butterworth filter) is used to filter out high-frequency noise in the tension signal (such as noise caused by high-frequency wave impact or sensor electronic noise). The cutoff frequency of the filter can be adaptively adjusted. Specifically, the main motion frequencies of the platform (such as pitch and roll frequencies) are estimated based on real-time wind and wave data, and the cutoff frequency is dynamically set to be higher than the overall motion frequency of the platform to retain effective load information, while being lower than the local vibration frequency of the mooring cable to filter out interference.

[0032] The aforementioned adaptive filtering effectively preserves low-frequency tension fluctuations that reflect the overall mechanical state of the mooring system, while accurately removing irrelevant noise and improving the data signal-to-noise ratio.

[0033] Meanwhile, to eliminate the time difference between data acquisition and transmission from different sensors, a unified timestamp is applied to each set of data acquired simultaneously (tension data and wind and wave direction data) to eliminate the data asynchrony problem caused by transmission delay.

[0034] Step S120: Based on the platform structure of the floating platform and the layout information of each mooring point, construct a tension distribution topology map and map the tension data of each mooring point to the tension distribution topology map.

[0035] In practice, the positions of each mooring point on the floating platform are obtained; Each mooring point is defined as a node, and based on the position of each mooring point on the floating platform, the nodes corresponding to spatially adjacent mooring points that have a direct mechanical coupling relationship through the platform structure are connected by edges to construct a static initial tension distribution topology. The tension data collected in real time at each mooring point is used as the node feature of the corresponding node and assigned a value to obtain a tension distribution topology map that reflects the tension transmission path of the mooring system in physical space.

[0036] The implementation process of this step will be described in detail below: First, construct the graph structure of the static topology graph, including: 1) Node definition: Each mooring point is abstracted as a node in the graph. For example, a system with 6 mooring points has 6 nodes in its topology graph.

[0037] 2) Edge Definition: Based on the platform's geometry and mechanical properties, determine the connection relationships between nodes, i.e., edges. The specific rule is: analyze the platform structure and connect the nodes corresponding to spatially adjacent mooring points that have a direct force transmission path through the platform's main structure (such as beams, frames, etc.) with edges. For example, on a hexagonal platform, the two corner points at both ends of each edge are adjacent and directly connected by the frame; therefore, an edge should be established between their corresponding nodes.

[0038] Finally, based on the nodes and edges defined above, a static initial tension distribution topology G(V,E) reflecting the inherent structure of the floating photovoltaic platform is constructed, where V is the set of nodes and E is the set of edges. Then, the tension data collected in real time at each mooring point is assigned as the node features of the corresponding node to obtain a dynamic tension distribution topology reflecting the tension transmission path of the mooring system in physical space.

[0039] Furthermore, the dynamic tension distribution topology obtained in this application adopts the most direct method of taking the real-time tension data (if there is filtering, the filtered real-time tension data here) corresponding to each mooring point as the node feature of its corresponding node in the static initial tension distribution topology. That is, the real-time tension data Ti of each mooring point is assigned as the node feature of the corresponding node Node_i.

[0040] To further enhance feature representation capabilities, node features can be expanded from a scalar (i.e., the tension data of the corresponding node) into a multidimensional feature vector. This vector may include: 1) Real-time tension data T_instant; 2) The slope of the tension change trend, T_trend, calculated based on historical data over a past period (e.g., the last 10 minutes), can be obtained through linear fitting. A negative slope indicates that the anchorage is slowly failing.

[0041] 3) The statistical variance T_variance of the current tension data within the slip time window (e.g., the most recent 1 minute). High variance may indicate intermittent slippage or irregular impact on the mooring cable.

[0042] Thus, each node carries a rich feature vector [T_instant, T_trend, T_variance], which simultaneously contains instantaneous, trend, and stability information of the tension.

[0043] The above-described fusion implementation allows this application to simultaneously sense the instantaneous state, trend of change, and stability of tension, thereby enabling the detection of complex anomalies such as slow anchoring failure (negative trend growth) and intermittent slippage of mooring cables (drastic increase in variance).

[0044] Furthermore, to ensure that the topology graph reflects the primary and secondary relationships of real-time load transfer, this application can assign dynamic weights to each edge. Specific implementations include: For each edge, calculate the angle α between the real-time wind and wave direction θ_env and the normal direction of the platform structure segment represented by that edge. The smaller the angle α, the more directly the structure segment faces the wind and waves, and the more directly the load is transferred, i.e., it is the main load transfer path. Therefore, assign a higher weight coefficient to this edge, such as weight coefficient W=cos(α). When the angle is 0° (facing the wind and waves), the weight is 1; when the angle is 90° (parallel), the weight is 0.

[0045] Thus, a weighted, dynamic tension distribution topology graph G(V,E,W) is constructed, where V is the set of nodes, E is the set of edges, and W is the set of weights.

[0046] Step S130: Input the mapped tension distribution topology map into the pre-trained anomaly recognition model to identify abnormal nodes with abnormal tension distribution.

[0047] This step utilizes a pre-trained model to analyze the constructed dynamic tension distribution topology to identify anomalies. This application provides two optional implementation methods, including: Method 1: The pre-trained anomaly detection model is constructed using a graph neural network (GNN). Step 1: Input the dynamic tension distribution topology graph with node features (or node feature vectors) into the pre-trained graph neural network (GNN) model. Step 2: Through at least one layer of graph convolution operation in the graph neural network, perform the following operations for any layer: aggregate the tension features of each node itself and its neighboring nodes to learn the spatial dependencies between nodes in the tension distribution topology graph; specifically: aggregate the feature information of each node itself and its neighboring nodes (nodes connected by edges). When edges have weights, higher attention is given to the features of nodes connected by high-weight edges during aggregation. Through multiple stacked graph convolutions, each node can perceive information from distant nodes in the graph. Thus, after at least one layer of graph convolution operation is performed, through the above aggregation, the GNN model can learn the normal proportion and coordination relationship (i.e., spatial dependency) between the tension data of each mooring point under complex platform structures when environmental loads come from a certain direction. For example, it learns that when there is a northerly wind, the tension ratio of node 1 and node 2 should generally be within a certain range.

[0048] Step 3: Based on the learned spatial dependencies, output the confidence score for each node indicating abnormal tension, and identify anomalous nodes based on the confidence score for each node. Specifically: The output layer of the GNN model typically outputs a confidence score (between 0 and 1) for each node to represent its degree of anomalousness. A threshold (e.g., 0.7) is set, and nodes with confidence scores higher than this threshold are identified as anomalous nodes.

[0049] It is evident that the GNN method can automatically learn extremely complex normal tension distribution patterns end-to-end and has a strong detection capability for anomalies that are difficult to describe with explicit rules (such as complex faults where multiple anchor points influence each other).

[0050] Method 2: The pre-trained anomaly detection model is constructed through residual analysis; Step 1: Based on the real-time wind and wave direction data θ_env, using a pre-established physical rule model or empirical lookup table, predict the theoretically expected normal tension data T_expected_i for each node in the tension distribution topology map under the current environmental load. Use these predicted values ​​to construct a baseline tension distribution topology map G_expected.

[0051] Step 2: Calculate the residual between the real-time tension distribution topology map composed of real-time tension data and the reference tension distribution topology map; specifically, compare the real-time tension distribution topology map G_real with the reference tension distribution topology map G_expected node by node, and calculate the residual value of each node: Residual_i=|T_real_i-T_expected_i|.

[0052] Step 3: Identify nodes whose residual value Residual_i exceeds a preset threshold as abnormal nodes. This preset threshold can be set for different regions (e.g., high-tension zones, low-tension zones).

[0053] It is evident that the principle of residual analysis is intuitive, computationally efficient, and particularly suitable for scenarios with relatively clear mechanical models, making it easy to understand and verify.

[0054] Step S140: Based on the location of the mooring point corresponding to the abnormal node and the real-time wind and wave direction data, locate the failed target seabed anchor.

[0055] Based on the real-time wind and wave direction data θ_env, the main force direction of the floating platform is determined, and the expected high-tension mooring point area (windward side) and the expected low-tension mooring point area / unexpected high-tension mooring point area (leeward side) are divided accordingly. Next, the location of the abnormal node is compared with the expected high-tension mooring point area and the expected low-tension area mentioned above: (1) If an abnormal node is located in the expected high-tension mooring point area, but its tension data is abnormally low, it indicates that the seabed anchor corresponding to the node has failed (e.g., the anchor has been dragged out of the seabed) and cannot provide sufficient tensile strength.

[0056] (2) If the abnormal node is located in the expected low-tension mooring point area / unexpected high-tension mooring point area, but its tension data is abnormally high, it may be due to the failure of other anchor points causing load redistribution, forcing the mooring cable, which should be relaxed, to bear additional tension, thus forcing the node to bear additional load. In this case, combined with the topology, the most likely primary failure of the seabed anchor can be inferred in reverse.

[0057] In some embodiments, to minimize false alarms, after initially determining the target seabed anchor based on tension logic, this application can also initiate additional monitoring for verification. For example, when a seabed anchor is suspected of being faulty, data from underwater acoustic detectors at the corresponding location will be retrieved to analyze whether there are abnormal sounds of mooring cables dragging or striking the seabed; simultaneously, platform tilt data will be analyzed to see if there are any minute attitude changes consistent with the abnormal tension pattern. The failure of the target seabed anchor is finally confirmed only when the tension anomaly and at least one additional monitoring signal anomaly are present simultaneously.

[0058] Specifically: When a seabed anchor (such as A3) is preliminarily determined to be faulty, the system automatically retrieves auxiliary sensor data related to that anchor point for the following verification: (1) Acoustic verification: retrieve data from underwater acoustic detectors deployed near anchor point A3 and analyze whether there are abnormal acoustic signals of specific frequencies related to mooring cable dragging, seabed impact, or structural damage during abnormal time periods.

[0059] (2) Attitude verification: retrieve the platform tilt sensor data and check whether the overall attitude of the platform has changed in a way that matches the slight tilt pattern that should be caused by the failure of the A3 anchor point.

[0060] The A3 seabed anchor is definitively confirmed as faulty only if at least one of the following conditions is met: abnormal tension, abnormal acoustic signal, or abnormal attitude. If only abnormal tension is present, and the auxiliary sensors show no abnormalities, it can be marked as a state to be observed, rather than immediately triggering an alarm.

[0061] In some embodiments, after confirming the failure of a seabed anchor, detailed alarm information can be generated and pushed to maintenance personnel through the monitoring center's large screen, web interface, or mobile application.

[0062] The alarm message includes a target identifier, anomaly information, and recommended actions. The target identifier can be the number of the failed seabed anchor (e.g., anchor A3). The anomaly information can include the anomaly type (e.g., insufficient anchoring force) and confidence level. Recommended actions can include specific maintenance suggestions (e.g., prioritizing visual inspection of anchor A3 using an underwater robot).

[0063] In some embodiments, the training process for the anomaly detection model is as follows: A large amount of historical data (tension, wind and wave direction) under normal operating conditions and various simulated fault conditions was collected. Simulated fault data can be obtained by simulating specific anchor point failures using a high-fidelity simulation model, or by injecting synthetic anomaly data conforming to mechanical laws into real historical data. In particular, adversarial operating condition samples can be introduced to simulate fault data, such as simulating complex situations where non-primary load anchors fail but are compensated for by the structure, to improve model robustness.

[0064] Supervised learning is employed to label the data (e.g., normal / abnormal, or abnormal node locations) in order to minimize the loss function between the model's predicted output and the true label.

[0065] Furthermore, the model can be continuously optimized: incremental learning is used for continuous optimization.

[0066] The correct alarms and false alarms confirmed by operations and maintenance personnel, along with their corresponding data, are used as new training samples to update the model regularly.

[0067] Furthermore, in groups with multiple floating photovoltaic power plants, a federated learning paradigm can be adopted. Each power plant's model is trained locally using local data. Only the updated model parameters need to be encrypted and uploaded to the cloud for aggregation, generating a more powerful global model before being distributed to each power plant. The original data remains on the local platform, enabling a collective improvement in the performance of all participating power plant models.

[0068] Corresponding to the above method, this application also provides a tension anomaly detection device for an anchoring system on a floating photovoltaic platform, such as... Figure 2 As shown, the device includes: The acquisition unit 210 is used to acquire tension data at each mooring point and real-time wind and wave direction data of the floating platform in real time. Construction unit 220 is used to construct an initial tension distribution topology map based on the platform structure and mooring point layout information of the floating platform; Mapping unit 230 is used to map the tension data of each mooring point to the initial tension distribution topology map; The identification unit 240 is used to input the mapped tension distribution topology map into the pre-trained anomaly identification model to identify abnormal nodes with abnormal tension distribution. The positioning unit 250 is used to locate the failed target seabed anchor based on the location of the mooring point corresponding to the abnormal node and real-time wind and wave direction data.

[0069] The functions of each functional unit of the tension anomaly detection device for the anchoring system on the floating photovoltaic platform provided in the above embodiments of this application can be implemented through the above-described methods and steps. Therefore, the specific working process and beneficial effects of each unit in the tension anomaly detection device for the anchoring system on the floating photovoltaic platform provided in the embodiments of this application will not be repeated here.

[0070] This application also provides an electronic device, such as... Figure 3 As shown, it includes a processor 310, a communication interface 320, a memory 330, and a communication bus 340, wherein the processor 310, the communication interface 320, and the memory 330 communicate with each other through the communication bus 340.

[0071] Memory 330 is used to store computer programs; When the processor 310 executes the program stored in the memory 330, it performs the following steps: Real-time acquisition of tension data at each mooring point and real-time wind and wave direction data of the floating platform; Based on the platform structure and mooring point layout information of the floating platform, an initial tension distribution topology map is constructed, and the tension data of each mooring point is mapped to this initial tension distribution topology map; The mapped tension distribution topology is input into a pre-trained anomaly detection model to identify anomalous nodes with abnormal tension distributions. Based on the location of the mooring point corresponding to the abnormal node and the real-time wind and wave direction data, the failed target seabed anchor was located.

[0072] The communication bus mentioned above can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc. For ease of illustration, only one thick line is used to represent it in the diagram, but this does not mean that there is only one bus or one type of bus.

[0073] The communication interface is used for communication between the aforementioned electronic devices and other devices.

[0074] The memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.

[0075] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0076] The implementation methods and beneficial effects of the various components of the electronic device in the above embodiments for solving the problem can be found in [reference needed]. Figure 1 The steps in the illustrated embodiments are used to implement the electronic device. Therefore, the specific working process and beneficial effects of the electronic device provided in this application will not be repeated here.

[0077] In another embodiment provided in this application, a computer-readable storage medium is also provided, which stores instructions that, when executed on a computer, cause the computer to perform the tension anomaly detection method for the anchoring system on the floating photovoltaic platform described in any of the above embodiments.

[0078] In another embodiment provided in this application, a computer program product containing instructions is also provided, which, when run on a computer, causes the computer to execute the tension anomaly detection method for the anchoring system on the floating photovoltaic platform described in any of the above embodiments.

[0079] Those skilled in the art will understand that the embodiments in this application can be provided as methods, systems, or computer program products. Therefore, the embodiments in this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the embodiments in this application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0080] This application describes embodiments of methods, apparatus (systems), and computer program products according to embodiments of this application with reference to flowchart illustrations and / or block diagrams. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0081] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0082] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0083] Although preferred embodiments have been described in this application, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the embodiments of this application.

[0084] Obviously, those skilled in the art can make various modifications and variations to the embodiments of this application without departing from the spirit and scope of the embodiments of this application. Therefore, if these modifications and variations to the embodiments of this application fall within the scope of the claims in this application and their equivalents, then this application also intends to include these modifications and variations.

Claims

1. A method for detecting abnormal tension in an anchoring system on a floating photovoltaic platform, characterized in that, The mooring system includes multiple mooring points symmetrically arranged on a floating platform and corresponding seabed anchors connected to each mooring point via mooring cables. The method includes: Real-time acquisition of tension data at each mooring point and real-time wind and wave direction data of the floating platform; Based on the platform structure and mooring point layout information of the floating platform, an initial tension distribution topology map is constructed, and the tension data of each mooring point is mapped to this initial tension distribution topology map; The mapped tension distribution topology is input into a pre-trained anomaly detection model to identify anomalous nodes with abnormal tension distributions. Based on the location of the mooring point corresponding to the abnormal node and the real-time wind and wave direction data, the failed target seabed anchor was located.

2. The method as described in claim 1, characterized in that, The tension data is measured by tension sensors installed at each mooring point; The real-time wind and wave direction data is measured by a wind vane and a wave vane installed on the floating platform.

3. The method as described in claim 1, characterized in that, Based on the platform structure and mooring point layout information of the floating platform, an initial tension distribution topology map is constructed, and the tension data of each mooring point is mapped to this initial tension distribution topology map, including: Obtain the position of each mooring point on the floating platform; Each mooring point is defined as a node, and based on the position of each mooring point on the floating platform, the nodes corresponding to spatially adjacent mooring points that have a direct mechanical coupling relationship through the platform structure are connected by edges to construct a static initial tension distribution topology. The tension data collected in real time at each mooring point is used as the node feature of the corresponding node and assigned a value to obtain a tension distribution topology map that reflects the tension transmission path of the mooring system in physical space.

4. The method as described in claim 1, characterized in that, The pre-trained anomaly detection model is built using a graph neural network; The mapped tension distribution topology is input into a pre-trained anomaly detection model to identify anomalous nodes with abnormal tension distributions, including: By aggregating the tension features of each node itself and its neighboring nodes through at least one layer of graph convolution operation in the graph neural network, the spatial dependencies between nodes in the tension distribution topology graph are learned. Based on the learned spatial dependencies, the confidence level of each node that there is an abnormal tension is output, and based on the confidence level of each node that there is an abnormal tension, the abnormal nodes are identified.

5. The method as described in claim 1, characterized in that, The pre-trained anomaly detection model is constructed through residual analysis; The mapped tension distribution topology is input into a pre-trained anomaly detection model to identify anomalous nodes with abnormal tension distributions, including: Based on the real-time wind and wave direction data, the expected tension data of each node in the tension distribution topology map under the current environmental load is predicted, and a baseline tension distribution topology map is generated. Calculate the residual between the real-time tension distribution topology map composed of real-time tension data and the reference tension distribution topology map; Nodes whose residual values ​​exceed a preset threshold are identified as abnormal nodes.

6. The method as described in claim 1, characterized in that, Based on the mooring point location and real-time wind and wave direction data corresponding to the anomaly node, the failed target seabed anchor was located, including: Based on the real-time wind and wave direction data, the main force direction of the floating platform and the corresponding expected high-tension mooring point area are determined. The location of the abnormal node is compared with the expected high-tension mooring point area; If the abnormal node is located in the expected high-tension mooring point area and the tension data of the abnormal node is abnormally low, then the seabed anchor corresponding to the abnormal node is determined to be the failed target seabed anchor.

7. The method as described in claim 1, characterized in that, After real-time acquisition of tension data at each mooring point and real-time wind and wave direction data of the floating platform, the method further includes: All tension data and real-time wind and wave direction data collected simultaneously in each group are unified under the same timestamp.

8. A tension anomaly detection device for an anchoring system on a floating photovoltaic platform, characterized in that, The mooring system includes multiple mooring points symmetrically arranged on a floating platform and corresponding seabed anchors connected to each mooring point via mooring cables. The device includes: The data acquisition unit is used to collect tension data at each mooring point and real-time wind and wave direction data of the floating platform. The building unit is used to construct an initial tension distribution topology map based on the platform structure and mooring point layout information of the floating platform; The mapping unit is used to map the tension data of each mooring point to the initial tension distribution topology map. The identification unit is used to input the mapped tension distribution topology map into the pre-trained anomaly identification model to identify abnormal nodes with abnormal tension distribution. The positioning unit is used to locate the failed target seabed anchor based on the location of the mooring point corresponding to the abnormal node and real-time wind and wave direction data.

9. An electronic device, characterized in that, The electronic device includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, used to store computer programs; A processor, when executing a program stored in memory, implements the method of any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method described in any one of claims 1-7.

Images