A dynamic cable dynamic response monitoring method based on deep learning, computer readable storage medium and program product

Abstract

A dynamic cable dynamic response monitoring method based on deep learning collects the six-degree-of-freedom motion time-history data of the floating platform and the three-degree-of-freedom motion time-history data of the buoy section through a monitoring system, and combines the dynamic response data of each monitoring node along the cable length to form a training data set for different fatigue hotspots. Then, a multi-task integrated model containing multiple parallel regression modules is constructed, and each module focuses on predicting the tension or bending moment of a specific monitoring node. Each regression module is trained separately using training data sets for different fatigue hotspots, and they are integrated to form a complete multi-task integrated model. Finally, the real-time collected motion time-history data is input into the model to predict the tension and bending moment of each monitoring node to realize the dynamic response monitoring of the dynamic cable. While realizing the health monitoring of the dynamic response of the dynamic cable operation and maintenance stage, the present invention reduces the monitoring cost and improves the economic benefits of the floating wind farm.

Description

Technical Field

[0001] The present invention relates to the intersecting fields of marine renewable energy, artificial intelligence and marine engineering technology, and in particular to a dynamic cable power response monitoring method based on deep learning. Background Art

[0002] Offshore wind power projects are booming, and the capacity of wind turbines is gradually increasing, the distance from the shore is increasing, and the development trend is gradually moving from nearshore to offshore. As the deep sea wind energy resources are more abundant, the wind energy quality is higher, and the impact on production and life in coastal areas is smaller, offshore wind power development has gradually tended to floating foundations in recent years to adapt to deeper waters. Dynamic cables are new components that appear in the stage of wind turbine systems turning from fixed foundations to floating foundations. They are of a gentle wave configuration. The upper part is connected to the floating platform, and the lower part is connected to the underwater booster station or buried in the shallow soil of the seabed. Buoyancy blocks are provided along the arc length to reduce the tension at the upper end of the dynamic cable. Dynamic cables not only undertake the tasks of power transmission and communication control, but also need to withstand external environmental loads such as waves and currents. In view of their functional characteristics and working characteristics, the safety of dynamic cables in the operation and maintenance stage of floating wind turbine systems has received widespread attention.

[0003] The typical three-core cross section of the dynamic cable is shown in Figure 1. The structural components of the dynamic cable from the inside to the outside are conductor 1, insulation layer 2, optical fiber 3, filler 4, inner sheath 5, armor 6 and outer sheath 7. The composite cross-section design makes the mechanical properties of the dynamic cable show nonlinear stress-strain relationship and stress coupling effect. The stress forms of the dynamic cable in the marine environment mainly include tension, bending moment and torsion. The armor layer with reverse winding direction is used in the design of the dynamic cable, which reduces the influence of torsion on the cable performance to a certain extent, so that tension and bending moment become the main internal force forms in the cross section of the dynamic cable. Figure 2 shows the stress distribution of a typical slow-wave configuration cable. There are stress peaks in the bottoming area and the suspension point of the cable, and multiple stress extremes appear near the buoy section, and the stress extremes at different positions are relatively close. The stress distribution of the dynamic cable with multiple extremes shows that the stress distribution of the slow-wave cable changes dramatically. Within the length range of the dynamic cable, there are multiple fatigue hot spots that control the safety of the dynamic cable. Therefore, the dynamic response monitoring of dynamic cables cannot only focus on the stress state at a specific point, but needs to comprehensively consider the stress distribution along the length of the dynamic cable.

[0004] The engineering practice of dynamic cable dynamic response monitoring is mainly achieved through integrated optical fiber distributed strain and temperature (DST) technology. However, the engineering application of this technology faces many challenges. In terms of monitoring cost, firstly, DST technology measures the strain of dynamic cables through optical fibers. The measurement of strain requires the dynamic cable cross section to adopt a tightly packed design. This cross-sectional design requires the dynamic cable manufacturing process to follow strict standards, which increases the production cost of dynamic cables. Secondly, optical fibers cannot withstand excessive strain, which places additional requirements on the installation process of dynamic cables, increasing the installation cost. In terms of the reliability of monitoring technology, at present. Optical fiber failure has become the main cause of dynamic cable failures, and the high failure rate of physical monitoring has further led to cable losses becoming the main cause of insurance claims for offshore wind farms. In terms of maintenance difficulty, the maintenance of dynamic cables requires the cooperation of multiple equipment such as remote-controlled submersibles and maintenance and laying vessels, and involves multiple processes such as positioning and replacement. While increasing the maintenance cost, the wind turbine downtime caused by offshore operations is also very considerable, greatly reducing the operating efficiency and economic benefits of floating wind farms.

[0005] The shortcomings of current dynamic cable dynamic response monitoring technology in engineering applications have prompted researchers to explore the combination of deep learning as a potential solution for the full life cycle monitoring of floating wind turbine dynamic cables. The deep learning health monitoring solution can get rid of the reliance of dynamic cable monitoring on physical entities (such as optical fibers) while achieving the same goals as existing physical monitoring technologies, thus providing a more cost-effective option. In addition, the deep learning-based monitoring solution does not require additional downtime during installation, which is particularly suitable for upgrading the monitoring technology of existing floating wind farms to avoid the impact of technology updates on wind farm operations. The current research on the dynamic response monitoring model of flexible cables based on deep learning still has a large gap in the application of dynamic cables. First of all, dynamic cables are different from anchor chains. There are multiple internal force forms such as tension and bending moment in their cross-sections. The dynamic response monitoring of dynamic cables needs to cover internal forces such as tension and bending moment at the same time to calculate the accumulated fatigue damage of the cross-section in real time. Secondly, the monitoring range of the current deep learning-based flexible cable dynamic response monitoring model is limited to a certain most dangerous point, while the gentle wave configuration of the dynamic cable causes its stress distribution along the cable length to vary dramatically. It is also affected by multiple loads such as waves and currents, as well as the movement of upper floating wind turbines. Its stress variations are complex, and there are multiple fatigue hotspots, requiring dynamic response monitoring at multiple nodes along the cable length.

[0006] It should be noted that the information disclosed in the above background technology section is only used for understanding the background of the present application, and therefore may include information that does not constitute prior art known to ordinary technicians in the field. Summary of the invention

[0007] The main purpose of the present invention is to overcome the defects existing in the above-mentioned background technology and provide a dynamic cable dynamic response monitoring method based on deep learning.

[0008] To achieve the above object, the present invention adopts the following technical solutions:

[0009] A dynamic cable dynamic response monitoring method based on deep learning comprises the following steps:

[0010] S1. Collect the six-degree-of-freedom motion time history data of the floating platform and the three-degree-of-freedom motion time history data of the buoy section through the monitoring system, and combine the motion time history data of the floating platform and the buoy section with the dynamic response data monitored by each monitoring node along the cable length to form a training data set for different fatigue hot spots of the dynamic cable;

[0011] S2. Construct a multi-task integrated model containing multiple parallel regression modules, where each regression module is used to predict the dynamic response of different monitoring nodes, including tension or bending moment;

[0012] S3. Use training data sets for different fatigue hotspots to train each regression module separately, and integrate the trained regression modules to form a complete multi-task integrated model;

[0013] S4. The motion time history data of the floating platform and the buoy segment collected in real time are input into the multi-task integrated model, and the tension and bending moment of each monitoring node are predicted by the multi-task integrated model to monitor the dynamic response of the dynamic cable.

[0014] Further, the fatigue hot spots of the dynamic cable include one or more of a hanging point HOP, a touchdown point TDP, a sag bending point SBP, a camber bending point HBP, a buoy section inlet ENP and a buoy section outlet EXP.

[0015] Furthermore, the contact point between the cable and the seabed at an arc length of 331m is used as the monitoring node for tension and bending moment; the entry and exit points of the buoy section at arc lengths of 131m and 257m are used as monitoring nodes for tension; and the extreme points of the static bending moment at arc lengths of 110m and 161m are used as monitoring nodes for bending moment.

[0016] Further, the movement of the buoy segments is monitored by underwater sensors installed at points of the buoy segments close to the sea surface.

[0017] Furthermore, the underwater sensor is arranged at a position where the cable arc length is 158 m.

[0018] Furthermore, the regression module adopts a long short-term memory network LSTM structure.

[0019] Furthermore, the six degrees of freedom of the floating platform include three translational degrees of freedom and three rotational degrees of freedom, and the three degrees of freedom of the buoy section include three translational degrees of freedom.

[0020] Furthermore, the monitoring system includes a global navigation satellite system and / or a mechanical sensor.

[0021] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the deep learning-based dynamic cable dynamic response monitoring method.

[0022] A computer program product includes a computer program, which, when executed by a processor, implements the deep learning-based dynamic cable dynamic response monitoring method.

[0023] The present invention has the following beneficial effects:

[0024] Aiming at the pain points of dynamic cable physical monitoring and the gap between the current flexible cable mechanical response monitoring model and the dynamic cable mechanical response monitoring, the present invention provides a floating wind turbine dynamic cable mechanical monitoring solution based on deep learning, and designs a dynamic cable multi-task integrated model based on deep learning to realize the dynamic response monitoring of dynamic cables. The model can realize the prediction of different internal forces in the cross-section of the dynamic cable at different positions along the cable length, while realizing the healthy monitoring of the dynamic response of the dynamic cable during the operation and maintenance stage, reducing the monitoring cost and improving the economic benefits of the floating wind farm. The solution of the present invention can significantly reduce the hardware requirements of the current dynamic cable mechanical monitoring, and comprehensively reduce the multi-stage costs of dynamic cable design, construction, installation, operation and maintenance.

[0025] The dynamic cable dynamic response monitoring method of the present invention can predict and analyze the dynamic response of the dynamic cable in the floating wind farm. The method can monitor multiple fatigue hot spots at the same time by constructing a multi-task integrated model, thereby comprehensively covering the entire length range of the dynamic cable. Compared with the traditional physical sensor method, the present invention reduces the dependence on expensive hardware and reduces the cost of manufacturing, installation and operation and maintenance. In addition, by introducing the motion characteristics of the buoy segment as the model input, the present invention can more accurately capture the impact of the floating platform and the buoy segment movement on the cable dynamic response, and improve the accuracy and reliability of monitoring. This method is not only suitable for specific cable configurations, but also has a wide range of applicability and can adapt to different flexible cable structures, thereby providing a more economical, efficient and technologically advanced monitoring solution for offshore wind power projects. In this way, the present invention can significantly improve the operating efficiency and economic benefits of floating wind farms, while ensuring the safety and reliability of wind farms.

[0026] Other beneficial effects of the embodiments of the present invention will be further described below. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 This is a schematic diagram of a typical three-core cross-section of a dynamic cable.

[0028] Figure 2 This is a schematic diagram of typical stress distribution of dynamic cables.

[0029] Figure 3 This is a thermal diagram of the dynamic cable fatigue hot spot tension correlation according to an embodiment of the present invention.

[0030] Figure 4 This is a thermal diagram of the correlation between the dynamic cable fatigue hot spot bending moment according to an embodiment of the present invention.

[0031] Figure 5 This is a schematic diagram of the structure of a dynamic cable multi-task integrated model according to an embodiment of the present invention.

[0032] Figure 6 This is a distribution diagram along the cable length according to an embodiment of the present invention.

[0033] Figure 7 This is a distribution diagram of static tension of a dynamic cable along the cable length without environmental load according to an embodiment of the present invention.

[0034] Figure 8 This is a distribution diagram of the static bending moment of a dynamic cable along the cable length without environmental load according to an embodiment of the present invention. DETAILED DESCRIPTION

[0035] The following is a detailed description of the embodiments of the present invention. It should be emphasized that the following description is only exemplary and is not intended to limit the scope and application of the present invention.

[0036] It should be noted that when an element is referred to as being "fixed to" or "disposed on" another element, it can be directly on the other element or indirectly on the other element. When an element is referred to as being "connected to" another element, it can be directly connected to the other element or indirectly connected to the other element. In addition, connection can be used for fixing as well as for coupling or communication.

[0037] It should be understood that the orientation or position relationship indicated by terms such as "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside" and "outside" are based on the orientation or position relationship shown in the accompanying drawings, and are only for the convenience of describing the embodiments of the present invention and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention.

[0038] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of the embodiments of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0039] The embodiment of the present invention provides a method for monitoring dynamic cable dynamic response based on deep learning, comprising the following steps:

[0040] S1. The six-degree-of-freedom motion time-history data of the floating platform and the three-degree-of-freedom motion time-history data of the buoy section are collected through the monitoring system, and the motion time-history data of the floating platform and the buoy section are combined with the dynamic response data monitored by each monitoring node along the length of the cable to form a training data set for different fatigue hot spots of the dynamic cable. The monitoring system may include a global navigation satellite system (such as a global positioning system GPS) and / or a mechanical sensor. The six-degree-of-freedom motion time-history data of the floating platform may be collected through the global positioning system GPS. The three-degree-of-freedom motion time-history data of the buoy section may be collected by monitoring the motion of the buoy section through an underwater sensor installed at a point of the buoy section close to the sea level.

[0041] S2. Construct a multi-task ensemble model consisting of multiple parallel regression modules (see Figure 5 ), where each regression module is used to predict the dynamic response of different monitoring nodes, including tension or bending moment.

[0042] S3. Use training data sets for different fatigue hotspots to train each regression module separately, and integrate the trained regression modules to form a complete multi-task integrated model.

[0043] S4. The motion time history data of the floating platform and the buoy segment collected in real time are input into the multi-task integrated model, and the tension and bending moment of each monitoring node are predicted by the multi-task integrated model to monitor the dynamic response of the dynamic cable.

[0044] In some embodiments, the fatigue hot spots of the dynamic cable include one or more of a hanging point HOP, a touchdown point TDP, a sag bending point SBP, a camber bending point HBP, a buoy section inlet ENP, and a buoy section outlet EXP.

[0045] In some embodiments, the regression module adopts a long short-term memory network LSTM structure.

[0046] In some embodiments, the six degrees of freedom of the floating platform include three translational degrees of freedom and three rotational degrees of freedom, and the three degrees of freedom of the buoy segment include three translational degrees of freedom.

[0047] This invention provides an innovative solution for dynamic cable dynamic response monitoring in offshore wind power projects through deep learning technology. As offshore wind farms develop towards deep seas, floating wind power foundations have become a key technology, and dynamic cables, as important components connecting wind turbine systems and underwater booster stations, are crucial to the stable operation of the entire wind farm. Dynamic cables are not only responsible for the transmission of electricity and communications, but also have to withstand complex marine environmental loads such as waves and currents, so their mechanical properties show nonlinear stress-strain relationships and stress coupling effects.

[0048] In view of the limitations of existing monitoring technologies, such as high cost, low reliability and complex maintenance process, this paper proposes a multi-task integrated model based on deep learning. The model can predict the tension and bending moment of dynamic cables at different positions along their length, realize the simultaneous monitoring of multiple fatigue hot spots, and thus improve the comprehensiveness and accuracy of monitoring. By integrating the six-degree-of-freedom motion data of the floating platform and the three-degree-of-freedom motion data of the buoy section, the model can comprehensively consider the joint influence of the upper floating platform and the middle buoy section on the cable dynamic response.

[0049] The present invention utilizes deep learning networks, especially long short-term memory networks (LSTM), to learn and predict the dynamic response of dynamic cables. This method not only reduces the reliance on physical sensors and the hardware cost of monitoring, but also improves the reliability and economic benefits of the monitoring process. Compared with traditional monitoring technologies, the model of the present invention can realize the monitoring of dynamic cables throughout their life cycle without introducing additional downtime during installation, and is particularly suitable for upgrading monitoring technology for existing wind farms. At the same time, the present invention also proposes a new input feature, the motion feature of the buoy segment, which captures the impact of the buoy segment on the overall dynamic response of the cable and comprehensively considers the joint impact of the floating platform and the buoy segment on the dynamic response of the cable, thereby improving the accuracy of the model prediction.

[0050] In general, the present invention uses deep learning technology to provide a dynamic cable dynamic response monitoring method that is economical, efficient, reliable and widely applicable, providing strong support for the stable operation of offshore wind farms and the improvement of economic benefits.

[0051] The specific embodiments of the present invention are further described below.

[0052] Take the slow-waveform dynamic cable in the DeepCwind semi-submersible platform floating wind turbine system as an example.

[0053] The fatigue hot spots of dynamic cables mainly include: Hang-off point (HOP) and touchdown point (TDP), sag bend point (SBP) and hog bend point (HBP), as well as the entrance point (ENP) and exit point (EXP) of the buoy section. The Spearman rank correlation coefficient is used To evaluate the correlation between tension and bending moment at different fatigue hotspots, the corresponding tension and bending moment correlation heat maps are shown in Figures 3 and 4, respectively. The tension correlation of different fatigue hotspots shows that the correlation between the tension at the suspension point and the tension at other locations is weak. In contrast, the correlation of the tension at the other five locations is relatively strong. The moment correlation of different fatigue hotspots shows that the correlation between the bending moment at different locations is obviously weak compared with the tension. It can be concluded that the tension and bending moment at different locations of the dynamic cable are not strongly correlated. The stress of the dynamic cable is composed of multiple internal forces such as bending moment and tension, so the dynamic response correlation of each fatigue hotspot of the dynamic cable is not strong. In view of the situation where the dynamic response correlation of different monitoring nodes is not strong, a multi-task integrated model is used to realize the stress prediction of tension and bending moment of the fatigue hotspot of the dynamic cable. The model consists of several parallel regression modules, each of which focuses on predicting the tension or bending moment at a specific location. The model structure of the multi-task integrated model is shown in Figure 5. The training data set is divided into training data of different fatigue hotspots and regression modules are trained for different fatigue hotspots. Different training modules together constitute a multi-task integrated model. New data is input into the multi-task integrated model, and its corresponding prediction is the dynamic response output of the dynamic cable fatigue hotspot.

[0054] The weak correlation of the dynamic response of the fatigue hotspot of the dynamic cable determines the model structure of the task integration. Selecting appropriate input features to establish the topological relationship between the input features and the dynamic response of each fatigue hotspot of the dynamic cable becomes a key step in establishing a dynamic cable multi-task integration model. The dynamic response monitoring model of the flexible cable represented by the dynamic cable and anchor chain uses the global navigation satellite system such as the Global Positioning System (GPS) and the effective data mapping of the structure motion time history and dynamic response. In terms of data quality, the monitoring data quality provided by the global positioning system is more stable than the monitoring data of the mechanical sensor. At the theoretical level, in the floating system, the data mapping between the motion time history of the floating body and the dynamic response time history of the flexible cable is supported by theoretical derivation. Considering that the dynamic response of the dynamic cable is jointly affected by the upper floating platform and the middle buoy section, accurately capturing the motion of the buoy section is crucial for cable status monitoring. In view of this, in addition to the six-degree-of-freedom motion time history input of the floating platform itself, the embodiment of the present invention also introduces an underwater sensor in the middle of the buoy to provide the floating body motion time history input to reflect the motion characteristics of the buoy section. When selecting the underwater sensor position, the selected point can not only reflect the motion characteristics of the buoy section, but also its motion can reflect the information of the current sea conditions to the greatest extent. Based on the above two considerations, the environmental load is more obvious at the position close to the sea level, and the movement of the position has a more significant impact on the cable movement. In addition, considering that the movement of the position with stronger environmental loads will have a more significant impact on the front and rear sections of the cable, the position close to the sea level is determined as the specific position of the underwater sensor to provide its three-degree-of-freedom motion. In summary, the inputs of the multi-task integrated model of the dynamic cable dynamic response model are the six-degree-of-freedom motion time history input of the floating platform and the three-degree-of-freedom motion time history input of the underwater sensor position of the buoy section.

[0055] The specific steps to establish a multi-task integrated model of dynamic cable dynamic response based on deep learning are as follows:

[0056] Determine monitoring nodes

[0057] It is crucial to determine the specific monitoring points along the cable length. Figures 6, 7 and 8 show the distribution of the static z coordinate, tension and bending moment of the cable along the cable length when there is no environmental load. In addition to the suspension point, the touchdown point is also a focus of attention. Figure 6 shows that at the arc length of 331 m, the dynamic cable contacts the seabed, which is manifested as a turning point in tension and a peak in bending moment at the same position in Figures 7 and 8. Therefore, the tension and bending moment at the arc length of 331 m are included in the monitoring scope. In addition to the suspension points and touchdown points that usually correspond to the ends of the cable, the determination of other key points is also based on the changes in tension and bending moment along the cable. In terms of tension, Figure 7 shows that the static tension has obvious peaks at arc lengths of 131 m and 257 m, which just corresponds to the starting point of the buoy section from 131 m to 257 m. Therefore, arc lengths of 131 m and 257 m are also included in the monitoring scope. In terms of bending moment, Figure 8 shows that the static bending moment has two extreme points at arc lengths of 110 m and 161 m, so the dynamic responses at these two points also need attention. Therefore, the coordinates of the monitoring nodes corresponding to the dynamic cable fatigue hot spots are 3 m, 110 m, 131 m, 161 m, 257 m, and 331 m, respectively.

[0058] Locate underwater sensors in the buoy section

[0059] The location of the underwater sensor of the buoy section should be determined based on the relative position between the dynamic cable buoy section and the sea level. According to the static longitudinal coordinate distribution shown in Figure 6, the point with an arc length of 158 m is the position closest to the sea level. Therefore, this point is selected as the specific location of the underwater sensor to monitor the movement of the buoy section and provide additional three-degree-of-freedom motion input for the multi-task integrated model. By obtaining the motion data of this point in real time as the motion data input of the multi-task integrated model, the overall motion state of the cable can be captured more effectively, thereby realizing the prediction of the dynamic response of the dynamic cable by the multi-task integrated model.

[0060] Multi-task ensemble model training

[0061] The training of the multi-task integrated model mainly focuses on the establishment, training and testing of different regression modules. In order to illustrate the effective prediction of dynamic cable dynamic response by task integration and buoy segment motion characteristics, this example uses the long short-term memory network (LSTM), a neural network structure commonly used in time series prediction, to establish, train and test the regression module. The sea conditions used to train the regression module are the typical working sea conditions of the DeepCwind semi-submersible platform floating wind turbine system. The input of the multi-task integrated model is the six-degree-of-freedom motion time history input of the floating platform and the three-degree-of-freedom motion time history input of the underwater sensor of the buoy segment, and the output is the bending moment and tension at the monitoring node position.

[0062] In summary, the present invention proposes a dynamic cable dynamic response monitoring method based on deep learning, and designs a dynamic cable dynamic response multi-task integrated model based on deep learning. The dynamic cable dynamic response model uses different regression modules for corresponding integration of the dynamic responses of different monitoring nodes, thereby achieving simultaneous output of the dynamic responses of different nodes, and realizing mechanical monitoring of different nodes within the length range of the flexible cable component. The present invention introduces the motion time history of the dynamic cable buoy section as the input of the multi-task integrated model, thereby considering the joint influence of the upper floating platform and the buoy section movement on the dynamic cable along the cable length, especially the dynamic response of the position below the buoy section.

[0063] The introduction of the motion characteristics of the buoy section is to set up underwater sensors at specific positions of the buoy section to reflect the movement of the buoy section. The location of the underwater sensor takes into account its relative position to the sea level to fully reflect the impact of environmental loads such as waves and currents on the surface of the sea on the dynamic cable movement.

[0064] Compared with the prior art, the advantages of the present invention mainly include:

[0065] The present invention designs a dynamic response monitoring model for multiple fatigue hotspots of dynamic cables. Compared with the use of physical sensors for dynamic response monitoring, the use of a dynamic cable dynamic response multi-task integrated model based on deep learning can effectively reduce the production, installation, and operation and maintenance costs of dynamic cables, and improve the economic benefits of floating wind power. Compared with the existing flexible cable mechanical response monitoring model based on deep learning, this model breaks through the limitation of the existing model for dynamic response prediction for a specific point, realizes the dynamic response prediction of multiple monitoring nodes, and extends the coverage of flexible cable dynamic response monitoring. The present invention proposes to introduce the motion characteristics of the buoy section as the input of the dynamic cable dynamic response monitoring model, and realizes the capture of the dynamic response of the fatigue hotspots distributed along the cable length of the dynamic cable by the dynamic response monitoring model. The dynamic response of the dynamic cable is simultaneously affected by the upper floating platform and the middle buoy section, accurately capturing the movement of the buoy section and accurately monitoring the status of the cable.

[0066] An embodiment of the present invention further provides a storage medium for storing a computer program, which at least performs the above method when executed.

[0067] An embodiment of the present invention further provides a control device, comprising a processor and a storage medium for storing a computer program; wherein the processor is configured to execute at least the method described above when executing the computer program.

[0068] An embodiment of the present invention further provides a processor, wherein the processor executes a computer program and at least executes the method described above.

[0069] The storage medium can be implemented by any type of non-volatile storage device, or a combination thereof. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a magnetic random access memory (FRAM), a ferromagnetic random access memory (Flash Memory), a magnetic surface memory, an optical disc, or a compact disc read-only memory (CD-ROM); the magnetic surface memory can be a disk memory or a tape memory. The storage medium described in the embodiments of the present invention is intended to include, but is not limited to, these and any other suitable types of memory.

[0070] In the several embodiments provided by the present invention, it should be understood that the disclosed systems and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0071] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0072] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0073] A person skilled in the art can understand that: all or part of the steps of implementing the above method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiment; and the aforementioned storage medium includes: mobile storage devices, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), disks or optical disks, etc. Various media that can store program codes.

[0074] Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention can be essentially or partly reflected in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROM, RAM, magnetic disks or optical disks.

[0075] The methods disclosed in the several method embodiments provided by the present invention can be arbitrarily combined without conflict to obtain new method embodiments.

[0076] The features disclosed in several product embodiments provided by the present invention can be arbitrarily combined without conflict to obtain new product embodiments.

[0077] The features disclosed in several method or device embodiments provided by the present invention can be arbitrarily combined without conflict to obtain new method embodiments or device embodiments.

[0078] The above contents are further detailed descriptions of the present invention in combination with specific preferred embodiments, and it cannot be determined that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art of the present invention, several equivalent substitutions or obvious variations can be made without departing from the concept of the present invention, and the performance or use is the same, which should be regarded as belonging to the protection scope of the present invention.

Claims

1. A dynamic cable dynamic response monitoring method based on deep learning, characterized in that: The following steps are involved: S1. Collect the six-degree-of-freedom motion time history data of the floating platform and the three-degree-of-freedom motion time history data of the buoy section through the monitoring system, and combine the motion time history data of the floating platform and the buoy section with the dynamic response data monitored by each monitoring node along the cable length to form a training data set for different fatigue hot spots of the dynamic cable; S2. Construct a multi-task integrated model containing multiple parallel regression modules, where each regression module is used to predict the dynamic response of different monitoring nodes, including tension or bending moment; S3. Use training data sets for different fatigue hotspots to train each regression module separately, and integrate the trained regression modules to form a complete multi-task integrated model; S4. The motion time history data of the floating platform and the buoy segment collected in real time are input into the multi-task integrated model, and the tension and bending moment of each monitoring node are predicted by the multi-task integrated model to monitor the dynamic response of the dynamic cable.

2. The dynamic cable dynamic response monitoring method based on deep learning as claimed in claim 1, characterized in that: The fatigue hot spots of the dynamic cable include one or more of a hanging point HOP, a touchdown point TDP, a sag bending point SBP, a camber bending point HBP, a buoy section inlet ENP and a buoy section outlet EXP.

3. The dynamic cable dynamic response monitoring method based on deep learning as claimed in claim 2, characterized in that: The contact point between the cable and the seabed at an arc length of 331m is used as the monitoring node for tension and bending moment; the entry and exit points of the buoy section at arc lengths of 131m and 257m are used as monitoring nodes for tension; the extreme points of static bending moment at arc lengths of 110m and 161m are used as monitoring nodes for bending moment.

4. The method for monitoring dynamic cable dynamic response based on deep learning according to any one of claims 1 to 3, characterized in that: The movement of the buoy segments is monitored by underwater sensors installed at points on the buoy segments close to the sea surface.

5. The method for monitoring dynamic cable dynamic response based on deep learning as claimed in claim 4, characterized in that: The underwater sensor is arranged at a position where the cable arc length is 158 m.

6. The method for monitoring dynamic cable dynamic response based on deep learning according to any one of claims 1 to 3, characterized in that: The regression module adopts a long short-term memory network LSTM structure.

7. The method for monitoring dynamic cable dynamic response based on deep learning according to any one of claims 1 to 3, characterized in that: The six degrees of freedom of the floating platform include three translational degrees of freedom and three rotational degrees of freedom, and the three degrees of freedom of the buoy section include three translational degrees of freedom.

8. The method for monitoring dynamic cable dynamic response based on deep learning according to any one of claims 1 to 3, characterized in that: The monitoring system includes a global navigation satellite system and / or a force sensor.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the deep learning-based dynamic cable dynamic response monitoring method as described in any one of claims 1 to 8 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the deep learning-based dynamic cable dynamic response monitoring method as described in any one of claims 1 to 8 is implemented.

Images