Automatic mooring load prediction method and system based on wind wave big data

Abstract

This invention provides a method and system for automatic mooring load prediction based on big data on wind and waves. First, wind and wave action data from the target sea area are collected to construct a wind and wave action chain, forming a set of wind and wave action sequences. Next, a load transmission network is constructed based on the distribution of structural components of the mooring equipment, forming a load transmission model. Then, the wind and wave action chain is input into the load transmission network, and a dynamic response mapping model is established by combining historical data. Real-time wind and wave action data is acquired, converted into a real-time wind and wave action chain, and input into the dynamic response mapping model to obtain the real-time load transmission distribution result and determine the predicted mooring load value. Finally, a load regulation strategy is generated based on the predicted mooring load value and sent to the control module to achieve dynamic load adjustment, thereby improving the accuracy and real-time performance of mooring load prediction and ensuring the safety of marine engineering.

Description

Technical Field

[0001] This invention relates to the field of marine engineering technology, and more specifically, to a method and system for automatic prediction of mooring load based on big data of wind and waves. Background Technology

[0002] In the field of marine engineering, mooring equipment plays a crucial role in ensuring the safe and stable operation of offshore facilities (such as ships and offshore platforms). Mooring equipment needs to withstand various complex loads from the marine environment, among which wind and waves are key factors affecting mooring loads.

[0003] Traditionally, the prediction of mooring loads has relied primarily on empirical formulas and simple physical models. These methods are often based on idealized assumptions and fail to accurately reflect the complexity and dynamism of wind and wave action in real marine environments. For example, wind and waves in the real ocean exhibit randomness and uncertainty; parameters such as wind and wave intensity, direction, and frequency vary significantly at different times and locations. Traditional methods typically cannot adequately account for the impact of these time-varying characteristics on mooring loads.

[0004] Furthermore, most existing prediction methods do not delve into the load transfer relationships between the internal structural components of mooring equipment, simply treating it as a whole for load calculation, leading to significant discrepancies between the predicted results and the actual situation. Moreover, these methods lack effective utilization of real-time wind and wave data, failing to adjust predictions promptly based on real-time changes in the marine environment, and thus failing to meet the requirements of modern marine engineering for the accuracy and real-time performance of mooring load predictions. Summary of the Invention

[0005] In view of the aforementioned problems, and in conjunction with the first aspect of the present invention, embodiments of the present invention provide an automatic mooring load prediction method based on wind and wave big data, the method comprising:

[0006] Collect wind and wave action data in the target sea area, construct wind and wave action chains based on the wind and wave action data, and integrate different wind and wave action types and their action parameters according to time series to form a set of wind and wave action sequences with time-series correlation;

[0007] Based on the distribution of structural components and the load transfer relationship between components of the mooring equipment, a load transfer network is constructed, the load receiving attributes of each structural component and the load transfer path between components are labeled, and a load transfer model of the mooring equipment is formed.

[0008] The wind-wave action chain is input into the load conduction network. By combining historical wind-wave action chains and historical load conduction data, a dynamic response mapping model of wind-wave action and load conduction is established. This model is used to output the corresponding load conduction distribution results based on the input wind-wave action parameters.

[0009] The real-time wind and wave action data of the target sea area is acquired, the real-time wind and wave action data is converted into real-time wind and wave action chains, the real-time wind and wave action chains are input into a dynamic response mapping model of wind and wave action and load conduction, real-time load conduction distribution results are obtained through dynamic correlation operation, and a mooring load prediction value is determined based on the real-time load conduction distribution results.

[0010] According to the mooring load prediction value and load bearing properties of each structural component of the mooring equipment, a load regulation strategy is generated, and the load regulation strategy is sent to a control module of the mooring equipment to realize dynamic adjustment of the load.

[0011] In another aspect, the embodiment of the present application also provides a mooring load automatic prediction system based on wind and wave big data, characterized in that it comprises:

[0012] A processor and a machine readable storage medium for storing machine executable instructions of the processor, wherein the processor is configured to execute the machine executable instructions to perform the above-mentioned mooring load automatic prediction method based on wind and wave big data.

[0013] In another aspect, the embodiment of the present application also provides a computer program product, which comprises machine executable instructions stored in a computer readable storage medium, and a processor of a computer device reads the machine executable instructions from the computer readable storage medium, and the processor executes the machine executable instructions to make the computer device execute the above-mentioned mooring load automatic prediction method based on wind and wave big data.

[0014] Based on the above aspects, by collecting wind and wave action data of the target sea area and constructing wind and wave action chains, the load conduction network constructed in detail labels load receiving properties of each structural component of the mooring equipment and load conduction paths between components, and describes the load transmission mechanism in the equipment, so that the load prediction can be deepened to the internal structure level of the equipment, and the accuracy of the prediction is improved. The dynamic response mapping model of wind and wave action and load conduction, combined with historical data, can truly reflect the complex dynamic relationship between wind and wave action parameters and load conduction distribution, and realize accurate mapping from wind and wave action to mooring load. Real-time load conduction distribution results are obtained by inputting real-time wind and wave action data into the model to determine the mooring load prediction value, which can respond to real-time changes of the marine environment in time, and ensure the real-time of the prediction results. Finally, the load regulation strategy generated according to the prediction value can realize dynamic adjustment of the load of the mooring equipment, effectively improve the safety and reliability of the mooring equipment, and reduce the risk of equipment damage and safety accidents caused by abnormal load. BRIEF DESCRIPTION OF DRAWINGS

[0015] Figure 1is an execution flow schematic diagram of the mooring load automatic prediction method based on wind wave big data provided by an embodiment of the present application.

[0016] Figure 2 is a schematic diagram of exemplary hardware and software components of the mooring load automatic prediction system based on wind wave big data provided by an embodiment of the present application. DETAILED DESCRIPTION

[0017] The present application will be described in detail below with reference to the accompanying drawings, Figure 1 is a flow schematic diagram of the mooring load automatic prediction method based on wind wave big data provided by an embodiment of the present application, which will be described in detail below.

[0018] Step S110: Collect wind wave action data of the target sea area, construct a wind wave action chain based on the wind wave action data, integrate different wind wave action types and action parameters under each type according to time sequence, and form a wind wave action sequence set with time sequence correlation.

[0019] In this embodiment, this step is carried out for the mooring system of the offshore floating production storage and offloading device. First, the target sea area where the device is located needs to be comprehensively collected for wind wave action data, and related information is integrated by constructing a wind wave action chain according to time sequence.

[0020] Step S111: Deploy wind wave data collection equipment at different monitoring points in the target sea area, collect wind wave action data at each monitoring point, and the wind wave action data includes information related to the duration, coverage range and intensity of wind wave action.

[0021] In this embodiment, in the target sea area around the floating production storage and offloading device, a plurality of monitoring circle layers are set according to a radius gradient, and a plurality of monitoring points are uniformly distributed in each monitoring circle layer. Ultrasonic wind speed and direction instrument, wave sensor, pressure sensor and other wind wave data collection equipment are deployed at each monitoring point. The ultrasonic wind speed and direction instrument is used to collect data such as the duration and direction of wind action; the wave sensor is used to collect data such as the duration and coverage range (i.e. the sea area affected by wave propagation) of wave action; and the pressure sensor is used to collect pressure values on the surface of the sensor to reflect information related to the intensity of wind wave action. Each collection equipment continuously collects data at a preset fixed frequency, and transmits the collected raw data to the data processing center in real time.

[0022] Step S112: Classify the collected wind wave action data, divide the wind wave action types according to the action forms of wind wave, and different wind wave action types correspond to different action performance characteristics.

[0023] In this embodiment, after the data processing center receives the wind and wave action data transmitted by each monitoring point, the classification processing program is started. According to the performance form of wind and wave action, the wind and wave action types are divided into stable wind action, gust action, regular wave action, irregular wave action and wind-wave combined action. Among them, the performance characteristics of stable wind action are that the wind speed and direction remain relatively stable for a long time, and the fluctuation amplitude is small; the performance characteristics of gust action are that the wind speed suddenly increases in a short time, and then quickly falls back, and the wind direction may be accompanied by a small deviation; the performance characteristics of regular wave action are that the period and amplitude of the wave show regular changes; the performance characteristics of irregular wave action are that the period and amplitude of the wave have no obvious rules and randomly fluctuate; the performance characteristics of wind-wave combined action are that wind and wave act on the target area at the same time, and the action parameters of the two influence each other. In the classification process, by extracting the wind speed change rate, wind direction fluctuation amplitude, wave period stability and other characteristic parameters in each data, the characteristic threshold of each type of wind and wave action is compared to determine the wind and wave action type to which each data belongs.

[0024] Step S113: For each wind and wave action type, extract the action parameters under the action of the wind and wave action type. The action parameters include the action intensity change, action direction deviation and action frequency distribution of the wind and wave action type acting on different monitoring points.

[0025] In this embodiment, after the classification of wind and wave action data is completed, for each determined wind and wave action type, the extraction of action parameters is carried out respectively to fully master the action characteristics of the type of wind and wave in the target sea area.

[0026] Step S1131: For a single wind and wave action type, select the wind and wave action data records of all monitoring points under the action of the single wind and wave action type.

[0027] In this embodiment, taking the stable wind action type as an example, all wind and wave data records marked as stable wind action are selected from the classified database. These records cover the relevant data collected by all monitoring points in the target sea area under stable wind action, including the position information, collection time, wind speed, wind direction and other contents of each monitoring point.

[0028] Step S1132: The wind and wave action data records of each monitoring point are arranged in chronological order, and the action intensity value corresponding to each time node is extracted.

[0029] In this embodiment, each data record under the selected stable wind effect is grouped according to the monitoring point. For each group of data at each monitoring point, the data is sorted in chronological order of collection time to form a time series data of the stable wind effect at that monitoring point. Subsequently, the wind speed value corresponding to each collection time node is extracted from the time series data and used as the intensity value of the stable wind effect at that monitoring point at that time node.

[0030] Step S1133: Calculate the difference between the intensity values ​​of adjacent time nodes. This difference is used to determine the trend of the intensity change of this single wind and wave action type at the monitoring point based on its positive and negative changes and the magnitude of the changes.

[0031] In this embodiment, taking the steady-state wind intensity value sequence at a certain monitoring point as an example, the difference between the intensity values ​​of two adjacent time nodes is calculated sequentially. If the intensity value of the later time node is greater than that of the earlier one, the difference is positive, indicating that the intensity is increasing; if the difference is negative, it indicates that the intensity is decreasing. Simultaneously, the magnitude of the intensity change is judged based on the absolute value of the difference; the larger the absolute value, the more significant the change in intensity between adjacent time nodes. By analyzing the differences between all adjacent time nodes, the overall trend of the steady-state wind intensity at that monitoring point throughout the entire monitoring period can be determined, such as whether it is continuously stable, slowly increasing, or fluctuating downwards.

[0032] Step S1134: Extract the action direction records of each time node under the single wind and wave action type at each monitoring point, count the duration of the same action direction, and determine the action direction offset based on the change in duration.

[0033] In this embodiment, taking the effect of stable wind as an example, wind direction records for each monitoring point at each time node are extracted. The wind direction is divided into preset directional intervals, such as every 15 degrees. Then, the duration of the wind direction at the monitoring point within each directional interval is counted throughout the entire monitoring period. By comparing the changes in the duration of the same directional interval in different time periods, the shift in the direction of action is determined. For example, if the duration of a certain directional interval decreases significantly within a certain time period, while the duration of adjacent directional intervals increases accordingly, it indicates that the wind direction has shifted towards that adjacent direction.

[0034] Step S1135: Divide the wind and wave action data records of each monitoring point under the single wind and wave action type into the same preset time window, and count the number of times the single wind and wave action type occurs within each preset time window.

[0035] In this embodiment, a preset time window duration is set, for example, 1 hour. The steady-state wind effect data records for each monitoring point are divided according to this preset time window, that is, the entire monitoring period is divided into multiple consecutive time windows of equal duration. Then, the number of times the steady-state wind effect occurs within each time window is counted; the number of occurrences here refers to the number of steady-state wind effect data records collected within that time window.

[0036] Step S1136: Calculate the frequency of action of this single wind and wave type at the monitoring point based on the number of occurrences within each preset time window and the duration of the preset time window.

[0037] In this embodiment, for each monitoring point, the frequency of the steady-state wind effect within that time window is obtained by dividing the number of occurrences of the steady-state wind effect within each preset time window by the duration of that preset time window. For example, if the steady-state wind effect occurs 60 times within a 1-hour time window (assuming a sampling frequency of 1 time / minute), then the frequency of the steady-state wind effect is 60 times / hour.

[0038] Step S1137: Arrange the frequency of action of different monitoring points according to the spatial distribution of the monitoring points to form a frequency distribution map of the single wind and wave action type.

[0039] In this embodiment, the location information of all monitoring points within the target sea area is marked on a two-dimensional coordinate system, with the horizontal axis representing longitude and the vertical axis representing latitude. Then, the calculated stable wind frequency for each monitoring point is mapped to its position in the coordinate system, and through interpolation and other processing methods, a stable wind frequency distribution map covering the entire target sea area is generated. This stable wind frequency distribution map can intuitively display the frequency differences of stable winds acting at different spatial locations within the target sea area.

[0040] Step S1138: Integrate the intensity change trend, direction shift, and frequency distribution of each monitoring point to form the action parameters corresponding to the single wind and wave action type.

[0041] In this embodiment, the data on the intensity variation trend and direction shift of the steady wind effect at each monitoring point, along with the frequency distribution map of the steady wind effect across the entire target sea area, are integrated to form a set of action parameters corresponding to the type of steady wind effect. This set of action parameters includes intensity variation and direction shift information categorized by monitoring point, as well as frequency spectrum information reflecting spatial distribution characteristics, comprehensively describing the action characteristics of steady wind effects. For other wind and wave action types, such as gusts and regular waves, the corresponding action parameters are extracted following the same process described above.

[0042] Step S114: Sort the different wind and wave action types and corresponding action parameters of each monitoring point in chronological order, and add time-series annotations to the sorted wind and wave action types and action parameters. The annotation information includes the start and end times of each wind and wave action type.

[0043] In this embodiment, different wind and wave action types and their corresponding action parameters are collected from all monitoring points, and the data is uniformly sorted according to the order of collection time. After sorting, the wind and wave action type corresponding to each data point is time-series labeled to determine the start and end times of each wind and wave action type at that monitoring point. For example, if a monitoring point starts collecting stable wind action data at a certain time and continues until another time when the stable wind action data disappears and is replaced by other types of wind and wave action data, then the previous time is marked as the start time of the stable wind action, and the next time is marked as the end time.

[0044] Step S115: Based on the time series labeling results, integrate the wind and wave action types and parameters of each monitoring point within the same time interval to form the wind and wave action unit corresponding to that time interval.

[0045] In this embodiment, several continuous and non-overlapping time intervals are divided based on the start and end times of the effects marked at all monitoring points. For each time interval, the types of wind and wave effects present at all monitoring points within that time interval and their corresponding effect parameters are collected and integrated. During the integration process, parameters such as the intensity changes, direction shifts, and frequency distributions of each type of wind and wave effect are categorized and organized according to the monitoring points, forming a wind and wave effect unit that contains information on wind and wave effects across the entire sea area within that time interval. Each wind and wave effect unit corresponds to a specific time interval, completely recording the wind and wave effect status of the target sea area during that time period.

[0046] Step S116: Connect the wind and wave action units in sequence according to the time interval to form a wind and wave action chain. Each wind and wave action unit in the wind and wave action chain maintains temporal connection and correlation of action parameters with the adjacent wind and wave action units.

[0047] In this embodiment, the predefined time intervals are arranged chronologically, and the corresponding wind and wave action units are then sequentially connected to form a complete wind and wave action chain. During the chaining process, it is ensured that the end time of the preceding wind and wave action unit is perfectly aligned with the start time of the following unit, without any time gaps. Simultaneously, the correlation of the action parameters of adjacent wind and wave action units is checked. For example, if the preceding unit is a steady wind action and the following unit transitions to a gust action, it is necessary to confirm whether there is a reasonable transitional relationship between the starting parameters of the gust action and the ending parameters of the steady wind action, such as whether the wind speed exhibits any abrupt changes, to ensure that the wind and wave action chain accurately reflects the continuous change process of wind and wave action.

[0048] Step S117: Perform a temporal integrity check on the constructed wind and wave action chain, supplement the missing wind and wave action data for the time intervals, and form a set of wind and wave action sequences with temporal correlation. The parameter dimensions of each wind and wave action unit in the set of wind and wave action sequences with temporal correlation remain consistent.

[0049] In this embodiment, a temporal integrity check is performed on the constructed wind and wave action chain. By comparing the preset total monitoring period with the time range covered by the wind and wave action chain, it is possible to find any uncovered missing time intervals. If a missing time interval is found, the cause of the missing time interval is first analyzed. If the data loss is due to a failure of the acquisition equipment, the action parameters of the wind and wave action units adjacent to the missing time interval are retrieved. Combined with the historical wind and wave data change patterns of similar sea areas during the same period, wind and wave action data for the missing time interval is generated through data interpolation and fitting. Based on this, corresponding wind and wave action units are constructed and added to the wind and wave action chain. At the same time, the parameter dimensions of all wind and wave action units in the wind and wave action chain are uniformly verified to ensure that each unit contains parameters of the same dimensions, such as monitoring point identification, wind and wave action type, action intensity change, action direction offset, and action frequency distribution. Finally, a set of wind and wave action sequences with temporal correlation is formed.

[0050] Step S120: Based on the distribution of structural components of the mooring equipment and the load transmission relationship between components, construct a load transmission network, label the load receiving attributes of each structural component and the load transmission path between components, and form a load transmission model of the mooring equipment.

[0051] In this embodiment, the mooring system of a floating production storage and offloading (FPSO) unit is taken as the research object. The mooring system consists of multiple structural components. By analyzing the distribution of these components and the load transfer relationship between them, a load transfer network is constructed, and a load transfer model is formed.

[0052] Step S121: Disassemble the overall structure of the mooring equipment and determine the core structural components of the mooring equipment. The core structural components include the force-bearing components that are in direct contact with wind and waves, the load-transmitting components, and the load-bearing support components.

[0053] In this embodiment, the overall structure of the mooring equipment of the floating production storage and offloading (FPSO) unit is disassembled and analyzed. The load-bearing components directly in contact with wind and waves include mooring cables, anchor chains, and buoys, which directly bear the forces exerted by the wind and waves. Load-transferring components include cable connectors, guide pulleys, and winches, used to transfer the load borne by the load-bearing components to other components. Load-bearing support components include hull connection seats, anchor foundations, and support frames, used to support the load of the entire mooring system and transfer the load to the seabed or the main hull structure. Through analysis of the structural drawings of the mooring equipment and actual disassembly and observation, the specific location, quantity, and specifications of the aforementioned core structural components are clarified.

[0054] Step S122: Analyze the physical properties of each core structural component, and determine the load receiving properties of each core structural component based on the physical properties. The load receiving properties include the load types that the core structural component can receive, the upper limit of load reception, and the deformation characteristics after load reception.

[0055] In this embodiment, physical property analysis is performed on each identified core structural component. For example, for a mooring cable, its physical properties include the material's elastic modulus, cross-sectional area, density, and tensile strength; for an anchor chain, these include the diameter of the chain links, material strength, and surface treatment; and for a buoy, these include buoyancy, weight, and geometric dimensions. Based on these physical properties, the load-bearing properties of each core structural component are determined. The type of load that a mooring cable can accept is tensile load. Its upper limit for accepting load is determined by the tensile strength and cross-sectional area of ​​the material, that is, the upper limit for accepting load is equal to the product of the tensile strength and the cross-sectional area. The deformation characteristics after accepting load are that elastic deformation occurs under tensile force, and the degree of deformation is positively correlated with the magnitude of tensile force, which conforms to Hooke's Law. The type of load that an anchor chain can accept is also tensile load. The upper limit for accepting load is determined by the strength of the chain link material and the structural dimensions. After accepting load, a certain amount of plastic deformation will occur. When the tensile force exceeds a certain limit, the deformation is irreversible. The types of loads that a buoy can accept are buoyancy load and pressure load. The upper limit for accepting load is determined by the magnitude of its buoyancy and its structural strength. After accepting load, a certain amount of floating and sinking displacement will occur, that is, the deformation characteristics are that the position moves up and down.

[0056] Step S123: Through structural mechanics analysis, determine the connection method between each core structural component. The connection method includes fixed connection, movable connection and elastic connection. Different connection methods correspond to different load conduction characteristics.

[0057] In this embodiment, structural mechanics analysis software is used to model and analyze the core structural components of the mooring equipment to determine the connection methods between the components. For example, the mooring cable is connected to the hull connecting seat via a flange, which is a fixed connection. The load transmission characteristics of this connection method are that it can completely transfer the tension borne by the mooring cable to the hull connecting seat, with almost no load loss and directional deviation. The connection between the mooring cable and the guide pulley is a movable connection, and the mooring cable can slide on the pulley. The load transmission characteristics of this connection method are that it can change the direction of load transmission, but a small amount of load loss will occur due to friction during the transmission process. The mooring cable is connected to the buoy via an elastic joint, which is an elastic connection. Its load transmission characteristics are that while transmitting the load, it can buffer part of the load impact through elastic deformation, reducing the instantaneous intensity of load transmission.

[0058] Step S124: Based on the connection method and load conduction characteristics, determine the load conduction path between adjacent core structural components. The load conduction path is marked with the conduction direction from the starting core structural component to the receiving core structural component and the load loss during the conduction process.

[0059] In this embodiment, the load conduction path between adjacent components is determined one by one according to the connection method and corresponding load conduction characteristics between each core structural component, and the key information of the path is marked.

[0060] Step S1241: For a pair of adjacent core structural components, determine the core structural component that bears the direct action of wind and waves as the starting core structural component, and the core structural component that receives the load transmitted by the starting core structural component as the receiving core structural component.

[0061] In this embodiment, the mooring cable and the guide pulley are selected as adjacent core structural components for analysis. Since the mooring cable is in direct contact with the wind and waves and bears the direct load applied by the wind and waves, it is determined as the starting core structural component; the guide pulley receives the load transmitted from the mooring cable and changes the direction of load transmission, therefore it is determined as the receiving core structural component.

[0062] Step S1242: Analyze the load conduction characteristics corresponding to the connection methods of the starting core structural component and the receiving core structural component. The load conduction characteristics of the fixed connection are that the load is transmitted without directional offset. The load conduction characteristics of the movable connection are that the load can be transmitted along the connection axis. The load conduction characteristics of the elastic connection are that there is elastic buffering during the load transmission process.

[0063] In this embodiment, the mooring cable and the guide pulley are movably connected, and the corresponding load transmission characteristic is that the load can be transmitted along the connecting axis of the guide pulley. That is, the tensile load borne by the mooring cable can be transmitted to the guide pulley along the rotation axis, and the guide pulley can adapt to changes in the tensile direction of the mooring cable by rotating, ensuring that the load can be continuously transmitted.

[0064] Step S1243: Determine the specific transmission direction of the load from the starting core structural component to the receiving core structural component based on the load transmission characteristics. The transmission direction must be consistent with the connecting axis direction of the core structural component and the structural force direction of the core structural component.

[0065] In this embodiment, the specific transmission direction is determined based on the load transmission characteristics of the movable connection between the mooring cable and the guide pulley, combined with the direction of the connecting shaft of the guide pulley and the direction of its structural stress. The connecting shaft of the guide pulley is set horizontally, and its structural design also dictates a horizontal stress direction. Therefore, the transmission direction of the load from the mooring cable to the guide pulley is determined to be horizontal, consistent with the direction of the connecting shaft and the direction of the structural stress. This ensures efficient load transmission and avoids uneven structural stress due to inconsistent directions.

[0066] Step S1244: Select the historical load transfer data of the adjacent core structural component, and extract the output load value of the starting core structural component and the input load value of the receiving core structural component from the historical load transfer data.

[0067] In this embodiment, historical load transfer data of the mooring cable and guide pulley, a pair of adjacent components, are selected from the historical operation database of the mooring equipment. This historical data covers the load values ​​output by the mooring cable and received by the guide pulley under different wind and wave conditions during multiple past operating cycles. During the extraction process, the historical data needs to be filtered in chronological order to ensure that the wind and wave type, intensity, and other conditions corresponding to the selected data segments are similar to the current analysis scenario, thereby improving the reliability of subsequent analysis results. The extracted content specifically includes the tension value of the end of the mooring cable away from the guide pulley at each time point (i.e., the output load value), and the tension value at the contact point between the guide pulley and the mooring cable (i.e., the input load value).

[0068] Step S1245: Calculate the difference between the output load value of the initial core structure component and the input load value of the receiving core structure component. This difference represents the load loss value during the conduction process.

[0069] In this embodiment, for each set of extracted mooring cable output load values ​​and guide pulley input load values, the difference between the two is calculated. Since the load is lost due to friction and other factors during transmission, the input load value of the guide pulley is usually less than the output load value of the mooring cable. Therefore, the calculated difference is positive, representing the load loss during transmission from the mooring cable to the guide pulley. By calculating multiple sets of data, a series of corresponding load loss values ​​can be obtained, reflecting the load loss situation of the transmission path under different operating conditions.

[0070] Step S1246: Statistically analyze the loss values ​​of multiple sets of historical load transmission data, and analyze the correlation between the loss values ​​and the output load values ​​and transmission direction of the initial core structural components.

[0071] In this embodiment, the calculated load loss values ​​are integrated and statistically analyzed with the corresponding mooring cable output load values ​​and conduction direction data. By plotting scatter plots, the variation of load loss values ​​with the mooring cable output load values ​​is observed, and the differences in load loss values ​​under different conduction directions (such as slight deviations in the horizontal direction) are analyzed. For example, when the mooring cable output load value increases, does the load loss value exhibit a linear or non-linear growth trend? When there is a slight deviation between the conduction direction and the direction of the guide pulley connection shaft, will the load loss value increase accordingly? Through the above analysis, the intrinsic correlation between the loss values, output load values, and conduction direction is clarified.

[0072] Step S1247: Determine the load loss patterns under different output load values ​​and different conduction directions based on the correlation relationship.

[0073] In this embodiment, based on the above correlation analysis results, the load loss pattern of the load conduction path under different conditions is summarized. For example, when the output load value of the mooring cable is within a certain range, the load loss value is linearly positively correlated with the output load value, that is, the larger the output load value, the larger the loss value; when the deviation of the conduction direction from the preset horizontal direction is within a certain threshold, the load loss value remains basically stable; once the deviation exceeds the threshold, the loss value will increase significantly. The above patterns are recorded in the form of textual description and curve fitting.

[0074] Step S1248: Mark the conduction direction and load loss pattern on the load conduction path of the adjacent core structural component to form a load conduction path description.

[0075] In this embodiment, in the constructed load conduction network model, the connecting line segment between the mooring cable and the guide pulley (i.e., the load conduction path) is located. The determined horizontal conduction direction is marked on this line segment, along with a recorded description of the load loss pattern. For example, the path is marked with "Conduction direction: horizontal; Loss pattern: when the output load value is within range A, the loss is linearly positively correlated with the output load, and the correlation coefficient is B; when the conduction direction deviation exceeds C, the loss increase is D," forming a complete description of the load conduction path. The load conduction paths between other adjacent core structural components are determined and marked according to the above steps S1241 to S1248.

[0076] Step S125: Label the parameters of each load conduction path, including the conduction efficiency, conduction delay and load conversion ratio during the conduction process.

[0077] In this embodiment, after determining and describing the load conduction path, detailed parameter annotations are further performed on each path to more accurately reflect the load conduction characteristics. The conduction efficiency is calculated as the ratio of the input load value of the receiving core structural component to the output load value of the starting core structural component; the closer this ratio is to 1, the higher the conduction efficiency. The conduction delay is determined by recording the time difference between the load output from the starting core component and the input from the receiving core component; this time difference is affected by factors such as the distance between components and the connection method. The load conversion ratio is determined for paths with load type conversion (e.g., some paths may convert tensile loads into torque loads) by calculating the ratio of the converted load to the original load. The calculated conduction efficiency, conduction delay, and load conversion ratio parameters are then annotated onto the corresponding load conduction path to complete the path's parameter information.

[0078] Step S126: Position each core structural component according to its spatial distribution location, and mark the position coordinates of each core structural component and the relative distance between core structural components in the spatial coordinate system.

[0079] In this embodiment, a three-dimensional spatial coordinate system is established, with a fixed reference point (such as the center of the hull) of the floating production storage and offloading (FPSO) unit as the origin. The x-axis, y-axis, and z-axis represent different spatial directions. Using laser rangefinders, GPS positioning, and other methods, the spatial position of each core structural component relative to the origin is measured, and the three-dimensional coordinate values ​​of each component are obtained and marked in the coordinate system. Simultaneously, based on the coordinate values ​​of each component, the straight-line distance (relative distance) between any two core structural components is calculated and recorded in the component attribute information.

[0080] Step S127: Based on the location coordinates of the core structural components, the load conduction path, and the load conduction path parameters, construct the topology of the load conduction network. In the topology of the load conduction network, each node represents a core structural component, and the connection between nodes represents the load conduction path.

[0081] In this embodiment, a three-dimensional spatial coordinate system is used as the basis. The position coordinates of each core structural component are taken as the position of the topology node, and node symbols representing each core structural component are drawn in the coordinate system. According to the previously determined load conduction path, lines are drawn between the corresponding nodes, and each line represents a load conduction path. During the drawing process, it is necessary to ensure that the direction of the lines is consistent with the actual load conduction direction, and that the relative positional relationship between the nodes matches the actual spatial distribution of the core structural components. At the same time, the labeled load conduction path parameters (such as conduction efficiency, conduction delay, etc.) are associated with the corresponding lines, so that the topology not only reflects the connection relationship between components, but also reflects the specific parameter characteristics of load conduction, thereby constructing a complete load conduction network topology.

[0082] Step S128: Supplement the load receiving attributes and load transmission path parameters of each core structural component in the topology of the load transmission network to form a load transmission model of the mooring equipment, which is used to present the transmission process of load inside the mooring equipment.

[0083] In this embodiment, based on the constructed load transmission network topology, corresponding load receiving attribute information is added to each node representing a core structural component. This includes the types of loads that can be received, the upper limit of load reception, and the deformation characteristics after load reception. This information is attached to the corresponding node in the form of node attribute tags. Simultaneously, the accuracy of parameters such as transmission efficiency, transmission delay, and load loss patterns marked on each load transmission path is reconfirmed to ensure the completeness and accuracy of the parameter information. Through the above supplementation and improvement, a load transmission model is formed that comprehensively reflects the attributes of each core structural component of the mooring equipment, the connection relationships between components, and the load transmission characteristics. This load transmission model can dynamically display the entire process of load transmission from the stressed components to the supporting components when wind and waves act on the mooring equipment, including the load loss, delay, and transformation along each path.

[0084] Step S130: Input the wind and wave action chain into the load conduction network, and combine historical wind and wave action chains and historical load conduction data to establish a dynamic response mapping model of wind and wave action and load conduction, which is used to output the corresponding load conduction distribution results according to the input wind and wave action parameters.

[0085] In this embodiment, the wind and wave action chain in the previously constructed set of wind and wave action sequences with temporal correlation is input into the load conduction network. At the same time, the historical operating data of the mooring equipment is called. By analyzing the correspondence between historical wind and wave action and load conduction, a dynamic response mapping model is established. This dynamic response mapping model can accurately output the load conduction distribution results inside the mooring equipment based on the input real-time wind and wave action parameters.

[0086] Step S131: Collect historical wind and wave action data for the target sea area within a set time period, and construct historical wind and wave action chains, with each historical wind and wave action chain corresponding to a historical time period.

[0087] In this embodiment, a relatively long historical data collection period is set, such as the past five years. All historical wind and wave action data for the target sea area within this period are retrieved from the data storage center. This data includes information such as the type and parameters of wind and wave action at different monitoring points and time periods. Following the same process and standards as constructing real-time wind and wave action chains, the aforementioned historical wind and wave action data are classified, parameters are extracted, and time-series integrated to construct multiple historical wind and wave action chains. Each historical wind and wave action chain corresponds to a continuous historical time period, such as one month's worth of historical wind and wave action data, ensuring that each historical wind and wave action chain has complete temporal correlation and parameter consistency.

[0088] Step S132: Collect historical load transmission data of the mooring equipment during the aforementioned historical time period, including the load values ​​of each core structural component at different times and the load changes along the load transmission path.

[0089] In this embodiment, historical load transmission data for the same period is collected from the historical operation database of the mooring equipment based on the historical time period corresponding to the constructed historical wind and wave action chain. This data is acquired through monitoring devices such as load sensors and strain gauges installed on each core structural component. Specifically, it includes the load values ​​(e.g., tension, pressure) borne by each core structural component at different times, the rate of load change, and the changes in load input, output, and loss values ​​over time along each load transmission path. During the collection process, the data undergoes preliminary screening to remove abnormal data caused by sensor malfunctions or other reasons, ensuring the reliability of the historical load transmission data.

[0090] Step S133: Associate each historical wind and wave action chain with the corresponding historical load transmission data to form a historical association dataset, and label the load values ​​of each core structural component corresponding to each historical wind and wave action unit.

[0091] In this embodiment, each historical wind and wave action chain is associated with historical load transmission data of the same period, using time as the link. Since each historical wind and wave action chain consists of multiple historical wind and wave action units, and each historical wind and wave action unit corresponds to a specific time interval, the historical load transmission data needs to be divided according to time intervals so that each historical wind and wave action unit can be mapped to the load values ​​of each core structural component within that time interval. During the association process, the average load value, maximum load value, and minimum load value of each core structural component within that time interval are labeled for each historical wind and wave action unit, forming a historical association dataset containing the correspondence between wind and wave action and load transmission.

[0092] Step S134: Extract the action parameters of the historical wind and wave action chain and the load transmission distribution characteristics from the historical correlation dataset. The load transmission distribution characteristics include the load ratio of each core structural component and the utilization rate of the load transmission path.

[0093] In this embodiment, feature extraction processing is performed on the historical associated dataset. The action parameters of each historical wind and wave action unit are extracted from each historical wind and wave action chain, such as changes in action intensity, shifts in action direction, and distribution of action frequency. Simultaneously, load transmission distribution features are extracted from the corresponding historical load transmission data. The load proportion of each core structural component is obtained by calculating the ratio of the average load value of a single core structural component to the sum of the average load values ​​of all core structural components. This ratio reflects the proportion of that component in the overall load borne by the mooring equipment. The utilization rate of the load transmission path is obtained by statistically analyzing the ratio of the actual load transmission time of that path per unit time to the total time; a higher ratio indicates a higher frequency of use and greater importance of the path. The extracted action parameters and load transmission distribution features are paired to form one-to-one feature pairs.

[0094] Step S135: Construct a mapping relationship training sample set, wherein each sample in the mapping relationship training sample set contains a set of wind and wave action parameters and corresponding load conduction distribution characteristics.

[0095] In this embodiment, each set of extracted wind and wave action parameters and corresponding load conduction distribution characteristics is used as a training sample. All the above training samples are summarized and organized to construct a mapping relationship training sample set. During the construction process, the sample set needs to be balanced to ensure that samples of different wind and wave action types and different load conduction distribution characteristics have a reasonable distribution ratio in the sample set, avoiding bias in subsequent model training due to uneven sample distribution. If it is found that the number of samples of certain types is too small, data augmentation methods are used, such as reasonably fine-tuning the parameters of existing similar samples to generate new samples, to supplement the number of samples, and finally form a mapping relationship training sample set of appropriate size and balanced distribution.

[0096] Step S136: Select a mapping relationship modeling method, and capture the dynamic correlation between the changes in wind and wave action parameters over time and the changes in load conduction distribution characteristics based on time-series correlation analysis.

[0097] In this embodiment, a Long Short-Term Memory (LSTM) network is selected as the mapping relationship modeling method. This method can effectively process time-series data and capture long-term dependencies between data. The LTM network receives time-series wind and wave impact parameters through its input layer. The information is then filtered and retained through the gating mechanism (input gate, forget gate, output gate) of the hidden layer, gradually capturing the patterns of wind and wave impact parameters changing over time, as well as the dynamic correlation between these changes and changes in load transmission distribution characteristics. For example, when the intensity of wind and wave impact shows a continuously increasing time-series change, the model can capture the corresponding changes in the load proportion of each core structural component, as well as the dynamic adjustment process of the utilization rate of different load transmission paths.

[0098] Step S137: Train the mapping relationship modeling method using the mapping relationship training sample set, and adjust the correlation parameters in the mapping relationship modeling process so that the output load transmission distribution characteristics are consistent with the actual characteristics in the mapping relationship training sample set.

[0099] In this embodiment, the selected long short-term memory network model is trained using the pre-constructed mapping relationship training sample set. By continuously adjusting the model's correlation parameters, the model's predictive performance is optimized.

[0100] Step S1371: Divide the mapping relationship training sample set into a training subset and a validation subset. The training subset is used for training the mapping relationship model parameters, and the validation subset is used for validating the mapping relationship model parameters.

[0101] In this embodiment, a random partitioning method is used to divide the mapping relationship training sample set into a training subset and a validation subset according to a preset ratio (e.g., 7:3). The training subset contains the majority of samples and is used for parameter learning and training of the model; the validation subset contains the remaining samples and is used to validate and adjust the model parameters during training to avoid overfitting. During the partitioning process, it is necessary to ensure that the distribution ratio of different types of samples in the training subset and the validation subset remains consistent with that of the original sample set to guarantee the effectiveness of training and validation.

[0102] Step S1372: Initialize the association parameters of the mapping relationship model, wherein the association parameters include the weight coefficient of the wind and wave action parameters, the influence coefficient of the load transmission path, and the attenuation coefficient of the time sequence association.

[0103] In this embodiment, the association parameters of the Long Short-Term Memory Network model are initialized. The weight coefficients of the wind and wave action parameters are used to measure the influence of different wind and wave action parameters (such as action intensity and action frequency) on the load transmission distribution characteristics. Initially, each parameter is assigned the same weight coefficient according to a uniform distribution. The influence coefficient of the load transmission path is used to reflect the importance of different load transmission paths in the load transmission process. Initially, different influence coefficients are preset according to the transmission efficiency of the path. The attenuation coefficient of the time series association is used to control the influence of historical time series data on the current prediction result. Initially, it is set to a fixed small value so that the model pays more attention to the influence of recent data.

[0104] Step S1373: Input the wind and wave action parameters of the first sample in the training subset into the mapping relationship model, and calculate and output the corresponding load transmission distribution prediction features based on the initial correlation parameters.

[0105] In this embodiment, the first sample is selected from the training subset. This sample contains a set of time-series wind and wave impact parameters. These wind and wave impact parameters are formatted according to the model input requirements and input into the initialized Long Short-Term Memory (LSTM) network model. Based on the preset initial correlation parameters, the model performs layer-by-layer calculations through the input layer and hidden layers, and finally outputs the corresponding load transmission distribution prediction features from the output layer. These load transmission distribution prediction features include the predicted load proportion of each core structural component and the predicted utilization rate of each load transmission path.

[0106] Step S1374: Extract the actual load conduction distribution features corresponding to the sample, calculate the degree of difference between the predicted load conduction distribution features and the actual load conduction distribution features, and measure the degree of difference by the sum of the numerical deviations of each dimension of the features.

[0107] In this embodiment, the actual load transmission distribution features corresponding to the input wind and wave action parameters are extracted from the training samples and compared with the predicted features output by the model. For each dimension of the features (such as the load proportion of each core structural component and the utilization rate of each path), the deviation between the predicted value and the actual value is calculated. Then, the deviation values ​​of all dimensions are added together to obtain the total degree of difference value. The larger this degree of difference value, the greater the deviation between the model's prediction result and the actual situation, and the worse the model performance.

[0108] Step S1375: Adjust the correlation parameters of the mapping relationship model according to the degree of difference. If the load ratio of any core structural component in the load transmission distribution prediction characteristics is higher than the actual load transmission distribution characteristics, reduce the weight coefficient of the wind and wave action parameter corresponding to the core structural component; if the predicted value of the load transmission path utilization rate is lower than the actual value, increase the influence coefficient of the load transmission path.

[0109] In this embodiment, the correlation parameters of the model are adjusted based on the calculated differences and their composition. For example, if the predicted load ratio of a core structural component is higher than the actual value, it indicates that the weighting coefficient of the wind and wave effect parameter corresponding to that component is too high, and the weighting coefficient needs to be appropriately reduced to decrease the prediction deviation. If the predicted utilization rate of a load transmission path is lower than the actual value, it indicates that the influence coefficient of that path is insufficient, and its influence coefficient needs to be increased so that the model can more fully consider the role of that path. At the same time, the attenuation coefficient of the time series correlation is adjusted according to the prediction deviation of the time series data. If historical data has a positive impact on the current prediction, the attenuation coefficient is appropriately increased, and vice versa.

[0110] Step S1376: Input the next sample in the training subset into the adjusted mapping model, and repeat the process of calculating the load transmission distribution prediction features, measuring the degree of difference, and adjusting the correlation parameters.

[0111] In this embodiment, the next sample in the training subset is selected, and its wind and wave action parameters are input into the model after parameter adjustment. The load conduction distribution prediction features are calculated again. Then, the degree of difference between the predicted features and the actual features is calculated using the same method as in step S1374. Based on the degree of difference, the correlation parameters of the model are further adjusted according to the adjustment rules in step S1375. Through the above sample-by-sample iterative method, the parameter settings of the model are continuously optimized.

[0112] Step S1377: After traversing all samples in the training subset, input the samples in the validation subset into the current mapping relationship model and calculate the overall degree of difference on the validation subset.

[0113] In this embodiment, after all samples in the training subset have completed one round of parameter tuning, all samples in the validation subset are input into the current model one by one to obtain the load propagation distribution prediction features for each sample. The degree of difference between the predicted features and the actual features of each validation sample is calculated, and then the average of the degree of difference of all validation samples is taken to obtain the overall degree of difference of the model on the validation subset, thereby evaluating the generalization performance of the model on data not involved in training.

[0114] Step S1378: If the overall difference is higher than the preset standard, continue to use the training subset for multiple rounds of parameter adjustment until the overall difference on the validation subset is lower than the preset standard.

[0115] In this embodiment, the overall difference level on the validation subset is compared with a preset standard threshold. If the overall difference level is higher than the preset standard threshold, the process returns to the training phase, and the model is re-tuned using the training subset for multiple rounds of parameter adjustments. In each new round of adjustments, the focus is on the sample types with large deviations in the previous validation round. The parameter adjustment strategy is refined based on the correlation between the wind and wave action parameters and the load conduction distribution characteristics corresponding to these samples. For example, if the prediction deviation of the load conduction distribution characteristics corresponding to a certain type of wind and wave action parameter is consistently large, the weight of this type of parameter in model training is appropriately increased to strengthen the model's learning of this type of data. After each round of parameter adjustments, the validation subset is re-inputted into the model to calculate the overall difference level, and this process is repeated until the overall difference level of the model on the validation subset is lower than the preset standard threshold, ensuring that the model has good generalization ability.

[0116] Step S1379: Record the final adjusted association parameters, fix the association parameters into the mapping relationship model, and complete the training process.

[0117] In this embodiment, once the overall difference of the model on the validation subset reaches a preset standard, parameter adjustment is stopped, and the final values ​​of all associated parameters in the model are recorded, including the weight coefficients of the wind and wave effect parameters, the influence coefficients of the load transmission path, and the attenuation coefficients of the time series correlation. These final associated parameters are then fixed to the corresponding parameter positions in the Long Short-Term Memory network model, generating a trained mapping model, thus completing the entire training process of the mapping model.

[0118] Step S138: Combine the trained mapping relationship model with the topology of the load transmission network, integrate the load receiving attributes and load transmission path parameters of each core structural component, and form a dynamic response mapping model of wind and wave action and load transmission.

[0119] In this embodiment, the trained mapping model is integrated with the previously constructed load conduction network topology. First, a data connection is established between the output layer of the mapping model and the nodes of the load conduction network topology, allowing the load conduction distribution characteristics output by the model to directly correspond to specific core structural components and load conduction paths within the topology. Then, the load receiving attributes (such as load type, receiving limit, and deformation characteristics) of each core structural component and the load conduction path parameters (such as conduction efficiency, conduction delay, and loss patterns) are integrated into the model's parameter system, enabling the model to fully consider the impact of component attributes and path characteristics on load conduction during calculation. Through this integration, a dynamic response mapping model for wind and wave action and load conduction is formed, possessing both time-series prediction and structural mapping capabilities. This dynamic response mapping model can output load conduction distribution results that conform to actual conduction patterns based on the input wind and wave action parameters and the structural characteristics of the mooring equipment.

[0120] Step S139: Input the test wind and wave action chain to the dynamic response mapping model of wind and wave action and load conduction, verify the consistency between the output load conduction distribution results and the actual load conduction data, and optimize the correlation parameters based on the verification results.

[0121] In this embodiment, a portion of historical wind and wave action data that was not used in model training and validation is selected, and a test wind and wave action chain is constructed using the same process. This test wind and wave action chain is input into the dynamic response mapping model, and the model outputs the corresponding load transmission distribution prediction result. Simultaneously, the actual load transmission data for the corresponding time period of the test wind and wave action chain is retrieved, and the prediction result is compared with the actual data one by one to calculate the overall deviation between the two. If the overall deviation is within an acceptable range, the model validation is successful; if the deviation exceeds the acceptable range, the cause of the deviation is analyzed, and the model's correlation parameters are fine-tuned accordingly, such as adjusting the parameter weights corresponding to specific wind and wave action types or the influence coefficients of specific load transmission paths. Validation is then performed again until the consistency between the prediction result and the actual data reaches the preset requirements, ensuring the accuracy and reliability of the dynamic response mapping model.

[0122] Step S140: Obtain real-time wind and wave action data of the target sea area, convert the real-time wind and wave action data into a real-time wind and wave action chain, input the real-time wind and wave action chain into the dynamic response mapping model of wind and wave action and load transmission, obtain the real-time load transmission distribution result through dynamic correlation calculation, and determine the mooring load prediction value based on the real-time load transmission distribution result.

[0123] In this embodiment, a real-time monitoring system for the target sea area is activated to continuously acquire current wind and wave action data. This data is then processed and transformed into a real-time wind and wave action chain that meets the model input requirements. The real-time load transmission distribution result is then calculated using a dynamic response mapping model, and finally, the mooring load prediction value is determined.

[0124] Step S141: Activate the real-time wind and wave monitoring equipment in the target sea area and acquire real-time wind and wave data at the preset acquisition frequency, including the type, intensity, direction and duration of wind and wave action at the current moment.

[0125] In this embodiment, real-time wind and wave monitoring equipment, including ultrasonic anemometers, wave sensors, and pressure sensors, deployed at various monitoring points in the target sea area, is activated. A preset data acquisition frequency is set, such as acquiring data once per minute. Each monitoring device continuously acquires wind and wave-related data at the current moment according to this frequency. Specifically, the ultrasonic anemometer acquires the current wind and wave type (e.g., steady wind, gust), direction (specific wind angle), and duration (the duration of this type of wind and wave); the wave sensor acquires the wave type (e.g., regular wave, irregular wave), intensity (wave height, wave speed), and duration; and the pressure sensor acquires the pressure value of the wind and waves acting on the sensor to assist in determining the intensity. All acquired real-time wind and wave data is transmitted to the data processing terminal in real time via a wireless communication module.

[0126] Step S142: Perform real-time classification of real-time wind and wave action data to determine the wind and wave action type corresponding to the current real-time wind and wave action data.

[0127] In this embodiment, after receiving real-time wind and wave action data, the data processing terminal activates the real-time classification module. This module calls preset wind and wave action type classification rules to extract characteristic parameters from the real-time data, such as wind speed change rate, wind direction fluctuation amplitude, and wave period stability. These characteristic parameters are then compared in real-time with characteristic thresholds for various wind and wave action types. For example, if the wind speed change rate in the real-time data is lower than a preset threshold and the wind direction fluctuation amplitude is small, the corresponding wind and wave action type is determined to be stable wind; if the wind speed change rate exceeds a preset threshold within a short period, it is determined to be a gust; if the wave period and wave height in the wave data show regular fluctuations, it is determined to be a regular wave. Through this real-time comparison, the wind and wave action type corresponding to each set of real-time wind and wave action data is quickly determined.

[0128] Step S143: Extract the action parameters from the current real-time wind and wave action data, including the real-time action intensity change, the real-time action direction offset, and the real-time action frequency distribution.

[0129] In this embodiment, after determining the real-time wind and wave action type, the corresponding action parameters are extracted for that type. For real-time action intensity changes, the intensity difference between adjacent acquisition times is calculated by continuously collecting action intensity values ​​(such as wind speed and wave height), and the upward or downward trend of intensity is analyzed. For real-time action direction shifts, the direction shift angle is calculated by comparing real-time wind direction or wave direction data with the direction at the previous acquisition time, and the shift trend is determined. For real-time action frequency distribution, the same type of wind and wave action data in the period before the current acquisition time is divided into preset time windows (such as 10 minutes), the number of occurrences of this type of wind and wave in each time window is counted, the action frequency is calculated by combining the time window duration, and a real-time action frequency distribution is formed according to the distribution of monitoring points.

[0130] Step S144: Connect the continuously collected real-time wind and wave action types and corresponding action parameters in chronological order to form a real-time wind and wave action unit. Each real-time wind and wave action unit corresponds to one collection cycle.

[0131] In this embodiment, the real-time wind and wave action types and extracted action parameters (real-time action intensity changes, direction shifts, and frequency distributions) determined within each acquisition cycle (e.g., 1 minute) are integrated to form a real-time wind and wave action unit. Each real-time wind and wave action unit includes the start and end times of the acquisition cycle, the wind and wave action type, and a complete set of action parameters, ensuring that each unit can fully reflect the wind and wave action status of the target sea area within the acquisition cycle.

[0132] Step S145: Integrate multiple consecutive real-time wind and wave action units in chronological order of acquisition time to construct a real-time wind and wave action chain.

[0133] In this embodiment, continuously generated real-time wind and wave action units are arranged sequentially according to their acquisition time to form a real-time wind and wave action chain. For example, real-time wind and wave action units from consecutive acquisition cycles such as 9:00-9:01 AM, 9:01-9:02 AM, and 9:02-9:03 AM are connected in chronological order to construct a real-time wind and wave action chain starting from 9:00 AM. During the construction process, it is ensured that the time connection between adjacent real-time wind and wave action units is seamless, and that the starting parameters of the later unit maintain a reasonable correlation with the ending parameters of the previous unit. For example, the trend of change in action intensity and the angle of direction offset can be continuously transitioned to truly reflect the dynamic change process of wind and wave action.

[0134] Step S146: Organize the real-time wind and wave action chain according to the input format required by the dynamic response mapping model of wind and wave action and load transmission, input the organized real-time wind and wave action chain into the dynamic response mapping model of wind and wave action and load transmission, call the internally trained mapping relationship model, and perform dynamic correlation calculations in combination with the topology and core structural component attributes of the load transmission network.

[0135] In this embodiment, the constructed real-time wind and wave action chain is formatted according to the preset input data format requirements of the dynamic response mapping model. For example, numerical data in the action parameters are converted into a standardized data format recognizable by the model, and each real-time wind and wave action unit is indexed and labeled according to time series. The formatted real-time wind and wave action chain is input into the dynamic response mapping model. The model automatically calls the internally trained Long Short-Term Memory (LSTM) network mapping relationship model, and simultaneously loads the topology data of the load transmission network and the load receiving attribute data of each core structural component, initiating dynamic correlation calculation. During the calculation, the mapping relationship model calculates the load changes of each core structural component and the load transmission process on each transmission path based on the temporal changes of the real-time wind and wave action parameters, combined with the component connection relationships and component attributes in the topology.

[0136] Step S147: During the calculation process, based on the parameters of each real-time wind and wave action unit in the real-time wind and wave action chain, calculate the load values ​​of each core structural component and the load changes of each load conduction path to obtain the real-time load conduction distribution results.

[0137] In this embodiment, during the dynamic correlation calculation stage, the corresponding load conduction distribution is calculated step by step for each real-time wind and wave action unit in the real-time wind and wave action chain, and finally integrated to obtain the complete real-time load conduction distribution result.

[0138] Step S1471: Extract the first real-time wind and wave action unit from the real-time wind and wave action chain, and obtain the wind and wave action type and action parameters corresponding to the first real-time wind and wave action unit.

[0139] In this embodiment, the first real-time wind and wave action unit is extracted from the constructed real-time wind and wave action chain in chronological order. This real-time wind and wave action unit corresponds to the earliest acquisition period. The corresponding wind and wave action type (such as steady wind) and complete action parameters are read from this real-time wind and wave action unit, including information such as the change in action intensity, action direction shift, and action frequency distribution of each monitoring point within the acquisition period.

[0140] Step S1472: Call the internal mapping relationship model, input the parameters of the first real-time wind and wave action unit into the mapping relationship model, and calculate the initial load values ​​of each core structural component under the action of the first real-time wind and wave action unit based on the trained correlation parameters.

[0141] In this embodiment, the long short-term memory network mapping relationship model within the dynamic response mapping model is invoked to input the action parameters of the first real-time wind and wave action unit into the model's input layer. Based on the correlation parameters determined during training (such as the weight coefficients of the wind and wave action parameters and the influence coefficients of the load transmission path), the model processes the parameters through the gating mechanism of the hidden layer. Combining this with the load receiving attributes of each core structural component (such as load type and receiving limit), the model calculates the initial load values ​​borne by each core structural component of the mooring equipment (such as mooring cables, guide pulleys, hull connecting seats, etc.) under the action of this wind and wave action unit. The initial load value of each core structural component corresponds to its acceptable load type (such as tensile load corresponding to tensile force value).

[0142] Step S1473: Based on the topology of the load conduction network, determine the load conduction path between each core structural component, and calculate the initial load conduction value from the starting core structural component to the receiving core structural component according to the conduction direction marked on the load conduction path and the load loss law.

[0143] In this embodiment, the load conduction path corresponding to the initial load is determined based on the connection relationships of each core structural component recorded in the load conduction network topology. For example, the mooring cable, as the initial core structural component, needs to have its initial load transferred to the guide pulley through the conduction path between it and the guide pulley. According to the horizontal conduction direction marked on this conduction path and the load loss law (such as the linear relationship between output load and loss), the initial load value of the mooring cable is subtracted from the loss value calculated according to the loss law to obtain the load conduction value transferred to the guide pulley. The guide pulley, as the receiving core structural component, receives the load and then continues to be transferred through the conduction path with the next component (such as the winch). Similarly, the conduction value is calculated according to the conduction direction and loss law of this path, and so on, to complete the conduction calculation of the initial load on each conduction path.

[0144] Step S1474: Record the changes in load conduction values ​​on each load conduction path, including the output load of the starting core structural component, the loss load in the load conduction path, and the input load of the receiving core structural component.

[0145] In this embodiment, while calculating the load transmission value, the load changes along each load transmission path are recorded in detail. For the transmission path from the mooring cable to the guide pulley, the output load value of the mooring cable (i.e., the initial load value), the loss load value along this path (calculated according to the loss law), and the input load value of the guide pulley (the difference between the output load and the loss load) are recorded. For the transmission path from the guide pulley to the winch, the output load value of the guide pulley (i.e., its input load value), the loss load value along this path, and the input load value of the winch are recorded. Through the above records, the details of load transmission along each transmission path are fully presented.

[0146] Step S1475: After completing the load calculation of the first real-time wind and wave action unit, extract the next real-time wind and wave action unit in the real-time wind and wave action chain, and input the parameters of the next real-time wind and wave action unit into the mapping relationship model.

[0147] In this embodiment, after completing the load conduction calculation and recording for the first real-time wind and wave action unit, the next real-time wind and wave action unit is extracted from the real-time wind and wave action chain in chronological order. This real-time wind and wave action unit corresponds to the next acquisition cycle. The wind and wave action type and action parameters of this real-time wind and wave action unit are input into the mapping relationship model to prepare for calculating the load change under the action of this wind and wave action unit.

[0148] Step S1476: Based on the final load values ​​of each core structural component under the action of the previous real-time wind and wave action unit, calculate the new load values ​​of each core structural component under the action of the current real-time wind and wave action unit.

[0149] In this embodiment, after receiving the parameters of the current real-time wind and wave action unit, the mapping relationship model first retrieves the final load values ​​of each core structural component under the action of the previous real-time wind and wave action unit (i.e., the load values ​​of each component at the end of the previous cycle). Based on the difference between the current wind and wave action parameters and the wind and wave action parameters of the previous cycle, and in conjunction with the correlation parameters, the newly added load value of each core structural component in the current cycle is calculated. For example, if the wind and wave action intensity in the current cycle is higher than that in the previous cycle, the newly added load value of the mooring cable is positive; if the intensity is lower, the newly added load value is negative.

[0150] Step S1477: Add the newly added load value to the final load value of the previous moment to obtain the real-time load value of each core structural component under the action of the current real-time wind and wave action unit.

[0151] In this embodiment, the calculated new load values ​​for each core structural component are superimposed with the final load values ​​at the end of the previous cycle to obtain the real-time load values ​​of each core structural component under the action of the current real-time wind and wave unit. For example, if the final load value of the mooring cable in the previous cycle is A, and the new load value in the current cycle is B, then the real-time load value of the mooring cable is the superposition of A and B. Simultaneously, according to the transmission direction and loss law of the load transmission path, the real-time load value of the current core structural component is transmitted to the next component, and the corresponding transmission value and path load change are calculated.

[0152] Step S1478: Repeat the above steps to complete the load calculation of all real-time wind and wave action units in the real-time wind and wave action chain in sequence, and obtain the load values ​​of each core structural component and the load changes of each load transmission path corresponding to each real-time wind and wave action unit.

[0153] In this embodiment, each real-time wind and wave action unit in the real-time wind and wave action chain is processed sequentially according to the process of steps S1475 to S1477. After processing each unit, the load values ​​of each core structural component corresponding to that unit and the load change data of each transmission path are obtained, until the load calculation of all real-time wind and wave action units is completed, forming a series of load calculation results arranged in chronological order.

[0154] Step S1479: Integrate the load calculation results corresponding to all real-time wind and wave action units in chronological order to form a real-time load transmission distribution result that includes the load values ​​of each core structural component at different times and the load changes of each load transmission path at different times.

[0155] In this embodiment, the load calculation results corresponding to each real-time wind and wave action unit are integrated in chronological order of acquisition time to construct a complete real-time load transmission distribution result. This real-time load transmission distribution result includes the real-time load values ​​of each core structural component at each acquisition time, forming a load time series for each component; it also includes the output load, loss load, and input load values ​​of each load transmission path at each acquisition time, forming a load change time series for each path. Through this result, the dynamic process of load transmission of mooring equipment under real-time wind and wave action can be comprehensively understood.

[0156] Step S148: Extract the real-time load values ​​of each core structural component from the real-time load transmission distribution results, and sum the load values ​​of all core structural components to obtain the total load value.

[0157] In this embodiment, the real-time load values ​​of each core structural component (such as mooring cables, anchor chains, guide pulleys, hull connectors, etc.) are extracted from the real-time load transmission distribution results, ensuring that the extracted values ​​are all from the same time and correspond to the same load type (e.g., tensile loads need to have unified units and types). The real-time load values ​​of all core structural components are summed to obtain the total load value of the mooring equipment at that time. This total load value reflects the total load scale currently borne by the mooring equipment.

[0158] Step S149: Determine the overall mooring load prediction value of the mooring equipment based on the sum of the load values ​​and the load proportion of each core structural component.

[0159] In this embodiment, the load percentage of each core structural component (i.e., the ratio of the load value of a single component to the sum of the load values) is extracted from the real-time load transmission distribution results. Combining the total load value and the changing trends of the load percentage of each component, a linear prediction method is used to predict the total load value in the next short period (e.g., the next 10 minutes). For example, if the current total load value is C, the load percentage of each component shows a stable trend in the near future, and the real-time wind and wave action parameters change gradually, then the predicted mooring load value for the next 10 minutes is C plus the incremental value calculated based on the rate of change of the wind and wave action parameters, ultimately determining the overall predicted mooring load value for the mooring equipment.

[0160] Step S150: Based on the predicted mooring load and the load-bearing attributes of each structural component of the mooring equipment, generate a load control strategy and send the load control strategy to the control module of the mooring equipment to achieve dynamic load adjustment.

[0161] In this embodiment, based on the determined mooring load prediction value and combined with the load-bearing capacity of each core structural component, a targeted load regulation strategy is formulated, and the regulation operation is executed by the control module to ensure the safe and stable operation of the mooring equipment.

[0162] For example, step S151: Extract the load-bearing attributes of each core structural component of the mooring equipment, including the maximum load-bearing value of each core structural component, the safe operating range of the load, and the warning threshold after the load exceeds the safe operating range of the load.

[0163] In this embodiment, load-bearing attribute data of each core structural component is extracted from the structural attribute database of the mooring equipment. For the mooring cable, its maximum load-bearing value is the product of the material's fracture strength and cross-sectional area. The safe operating range for the load is 0 to a certain percentage of the maximum load-bearing value (e.g., 80%), and the warning threshold is set to a certain percentage of the upper limit of the safe operating range (e.g., 90%). For the anchor chain, the maximum load-bearing value is determined by the chain link material strength and chain link diameter, usually calculated through structural mechanics. The safe operating range for the load is set to the same percentage as the maximum load-bearing value, and the warning threshold is also the corresponding percentage of the upper limit of this safe operating range. For the buoy, the maximum load-bearing value is divided into maximum buoyancy load and maximum pressure load. The maximum buoyancy load is determined by the buoy's displacement volume and seawater density, while the maximum pressure load is determined by the buoy shell material strength and thickness. The safe operating range for the load is set separately for each of the two load types, and the warning threshold is also determined according to the same percentage of the upper limit of the corresponding safe operating range. The extracted load-bearing attribute data of each core structural component is organized into a structured table and associated with a unique identifier for each component for easy subsequent querying and comparison.

[0164] Step S152: Based on the load proportion of each core structural component in the real-time load transmission distribution results, decompose the predicted mooring load value into the predicted load value of each core structural component.

[0165] In this embodiment, the current load percentage of each core structural component is obtained from the real-time load distribution results. This load percentage reflects the proportion of each component's share in the total load. Assuming the mooring load prediction is the total predicted load, the total predicted load is multiplied by the load percentage of each core structural component to obtain the predicted load value for each core structural component. For example, if the load percentage of the mooring cable is a certain proportion and the total predicted load is a certain value, then the predicted load value of the mooring cable is the product of the total predicted load and that proportion; the predicted load values ​​of other core structural components such as anchor chains and guide pulleys are calculated in the same way. After decomposition, the predicted load values ​​of each core structural component are associated and stored with the corresponding component identifier to form a component-predicted load lookup table.

[0166] Step S153: Compare the predicted load value of each core structural component with the load bearing attribute of the core structural component to determine the load status of each core structural component. The load status includes safe status, warning status and over-limit status.

[0167] In this embodiment, the predicted load values ​​of each core structural component in the component-predicted load comparison table are retrieved one by one and compared with the load-bearing attributes of that component extracted from the structural attribute database. If the predicted load value of a core structural component is less than or equal to the upper limit of its safe operating range and has not reached the warning threshold, the component is determined to be in a safe state; if the predicted load value exceeds the warning threshold but does not exceed the upper limit of the safe operating range, the component is determined to be in a warning state; if the predicted load value exceeds the upper limit of the safe operating range, the component is determined to be in an over-limit state. For example, the upper limit of the safe operating range of a mooring cable is a certain value, and the warning threshold is a corresponding proportion of this upper limit. If its predicted load value is between the warning threshold and the upper limit of the safe operating range, it is in a warning state; if it exceeds the upper limit of the safe operating range, it is in an over-limit state. The load status determination results of each core structural component, along with the component identifier and the predicted load value, are recorded to form a load status list.

[0168] Step S154: For core structural components in the early warning state, construct load mitigation measures, which include adjusting the spatial position of the core structural components to change the direction of force and enhancing the support strength of the core structural components to improve the load-bearing capacity.

[0169] In this embodiment, core structural components in a warning state are selected from the load status list. Based on their type, installation location, and load characteristics, targeted load mitigation measures are constructed. For load-bearing components such as mooring cables, if in a warning state, the length of the mooring cable can be adjusted by controlling the winch, changing the angle between the mooring cable and the hull, i.e., adjusting its spatial position, thereby changing the direction of force under wind and waves and dispersing some of the load. For transmission components such as guide pulleys, if a load warning is issued, temporary reinforcing ribs can be added to their support frame to enhance support strength, improve their load-bearing capacity, and prevent further load accumulation leading to deterioration. For each component in a warning state, specific operating parameters need to be defined, such as the length range of the mooring cable adjustment, the material and size of the reinforcing ribs, etc., to ensure that the mitigation measures are executable.

[0170] Step S155: For core structural components that are in an over-limit state, construct emergency load control measures, which include cutting off part of the load transmission path of the core structural component to reduce load input and activating backup support structures to share the load.

[0171] In this embodiment, for core structural components in an overload state on the load status list, higher-priority emergency load control measures need to be implemented to prevent component damage from causing safety accidents. If a mooring cable is in an overload state, the corresponding cable cutting device can be controlled to cut off part of the load transmission path between the mooring cable and other non-critical transmission components, reducing the load input to the mooring cable from external wind and waves. At the same time, a backup mooring cable near the mooring cable can be activated, using its support structure to share part of the load on the overloaded mooring cable, reducing its actual load value. For overloaded components such as anchor chains, backup fixing devices on the anchor foundation, such as backup anchor chains or hydraulic anchor piles, can be activated to share the load of the main anchor chain through the backup support structure, while simultaneously closing the connection between the anchor chain and unnecessary load transmission paths, reducing load input. The emergency load control measures need to clearly define the order of operation, such as activating the backup structure first and then cutting off part of the transmission path, to avoid sudden drops or increases in load during operation.

[0172] Step S156: After integrating the load control measures for each core structural component in the order of load transmission path, further determine the implementation timing of each load control measure based on the changing trend of the real-time wind and wave action chain, and construct the implementation steps of the load control measures, including the order of operation, the specific execution method of operation, and the monitoring indicators during the operation process.

[0173] In this embodiment, load mitigation measures and emergency load control measures for different core structural components are first integrated according to the order of load transmission paths in the load transmission network. For example, control measures for each component are arranged sequentially according to the transmission path sequence of "mooring cable → guide pulley → winch → hull connector" to ensure that the control measures of the previous link do not conflict with the next link. Subsequently, the changing trend of the real-time wind and wave action chain is analyzed. If the wind and wave intensity is increasing, load mitigation measures for components in the early warning state are implemented in advance. If the wind and wave intensity is already at a high level and remains stable, emergency control measures for components in the over-limit state are implemented immediately. Based on this, detailed implementation steps are constructed for each control measure: the order of operation is clarified, such as releasing the winch brake first, adjusting the length, and finally locking the brake when adjusting the position of the mooring cable; the specific execution method of the operation is specified, such as sending instructions to the drive motor of the winch through the control module to adjust the cable length at a preset rate; and monitoring indicators during the operation are set, such as real-time monitoring of the tension value of the mooring cable and the rotation speed of the guide pulley, to ensure that the operation process meets expectations.

[0174] Step S157: Organize the integrated load control measures, implementation timing and implementation steps into a load control strategy for the mooring equipment, and send the load control strategy of the mooring equipment to the control module of the mooring equipment to execute the control operation, so as to realize dynamic load adjustment.

[0175] In this embodiment, the integrated load control measures, determined implementation timings, and detailed implementation steps are organized into a standardized load control strategy document. This document includes the measure number, corresponding component identifier, implementation time window, detailed operation steps, and monitoring indicator thresholds. The load control strategy is sent to the mooring equipment's control module via a data transmission protocol. Upon receiving the document, the control module parses the strategy content and sends control commands to the corresponding actuators (such as winches, cutting devices, and backup structure activation devices) according to the implementation timing and operation steps. During execution, the control module receives real-time monitoring data from each core structural component and compares it with the monitoring indicators in the strategy. If the actual monitoring data deviates from expectations, the control commands are adjusted promptly to dynamically optimize the load control process until the load status of each core structural component is restored to a safe state, thus achieving dynamic adjustment of the mooring equipment's load.

[0176] Based on the same inventive concept, please refer to Figure 2 The diagram shows a schematic block diagram of a mooring load automatic prediction system 100 based on wind and wave big data, which is provided in an embodiment of this application for performing the above-described inspection video stream processing method. The mooring load automatic prediction system 100 based on wind and wave big data may include a communication unit 110, a machine-readable storage medium 120, and a processor 130.

[0177] In this embodiment, both the machine-readable storage medium 120 and the processor 130 are located within the mooring load automatic prediction system 100 based on wind and wave big data and are separately configured. However, it should be understood that the machine-readable storage medium 120 may also be independent of the mooring load automatic prediction system 100 based on wind and wave big data and may be accessed by the processor 130 via a bus interface. Alternatively, the machine-readable storage medium 120 may also be integrated into the processor 130 and may communicate and interact with external systems through the communication unit 110.

[0178] The processor 130 is the control center of the mooring load automatic prediction system 100 based on wind and wave big data. It connects to various parts of the system via various interfaces and lines. By running or executing software programs and / or modules stored in the machine-readable storage medium 120, and by calling data stored in the machine-readable storage medium 120, it performs various functions and processes data of the system, thereby providing overall monitoring of the system. Optionally, the processor 130 may include one or more processing cores; for example, the processor 130 may integrate an application processor and a modem processor, wherein the application processor mainly handles the operating system, user interface, and applications, and the modem processor mainly handles wireless communication. It is understood that the modem processor may also not be integrated into the processor. The machine-readable storage medium 120 is used to store machine-executable instructions for executing the scheme of this application, and the processor 130 is used to execute the machine-executable instructions stored in the machine-readable storage medium 120 to implement the inspection video stream processing method provided in the aforementioned method embodiments.

[0179] It should be noted that, in order to simplify the description of the present invention and thus help to understand one or more embodiments of the invention, multiple features may sometimes be grouped into one embodiment, drawing or description thereof in the foregoing description of the embodiments of the present invention.

Claims

1. A method for automatic prediction of mooring load based on big data of wind and waves, characterized in that, The method includes: Collect wind and wave action data for the target sea area, construct wind and wave action chains based on the data, and integrate different types of wind and wave action and the action parameters under each type of action according to the time series to form a set of wind and wave action sequences with time-series correlation. Based on the distribution of structural components of the mooring equipment and the load transfer relationship between components, a load conduction network is constructed, the load receiving attributes of each structural component and the load conduction path between components are labeled, and a load conduction model of the mooring equipment is formed. The wind-wave action chain is input into the load conduction network. By combining historical wind-wave action chains and historical load conduction data, a dynamic response mapping model of wind-wave action and load conduction is established. This model is used to output the corresponding load conduction distribution results based on the input wind-wave action parameters. Acquire real-time wind and wave action data for the target sea area, transform the real-time wind and wave action data into a real-time wind and wave action chain, input the real-time wind and wave action chain into a dynamic response mapping model of wind and wave action and load transmission, obtain the real-time load transmission distribution result through dynamic correlation calculation, and determine the mooring load prediction value based on the real-time load transmission distribution result. Based on the predicted mooring load and the load-bearing properties of each structural component of the mooring equipment, a load control strategy is generated and sent to the control module of the mooring equipment to achieve dynamic load adjustment. The data on wind and wave effects in the target sea area are collected. Based on this data, a wind and wave effect chain is constructed. Different types of wind and wave effects and their respective parameters are integrated in a time series to form a set of wind and wave effect sequences with temporal correlation, including: Deploy wind and wave data acquisition equipment at different monitoring points in the target sea area to collect wind and wave action data at each monitoring point. The wind and wave action data includes information related to the duration, coverage, and intensity of the wind and wave action. The collected wind and wave data are classified and processed, and the wind and wave action types are divided according to the manifestation of the wind and wave action. Different wind and wave action types correspond to different action characteristics. For each type of wind and wave action, the action parameters under that type of wind and wave action are extracted. The action parameters include the intensity change, direction shift and frequency distribution of the action of that type of wind and wave action at different monitoring points. The different wind and wave action types and corresponding action parameters of each monitoring point are sorted in chronological order. The sorted wind and wave action types and action parameters are time-series labeled, and the labeling information includes the start and end times of each wind and wave action type. Based on the time series labeling results, the wind and wave action types and parameters of each monitoring point within the same time interval are integrated to form the wind and wave action unit corresponding to that time interval. The wind and wave action units are connected in sequence according to the time interval to form a wind and wave action chain. Each wind and wave action unit in the chain maintains temporal connection and correlation of action parameters with the adjacent wind and wave action units. The constructed wind and wave action chain is checked for temporal integrity, and the missing wind and wave action data in the time intervals are supplemented to form a set of wind and wave action sequences with temporal correlation. The parameter dimensions of each wind and wave action unit in the set of wind and wave action sequences with temporal correlation remain consistent. Based on the distribution of structural components and the load transfer relationships between components of the mooring equipment, a load conduction network is constructed, and the load receiving attributes of each structural component and the load conduction paths between components are labeled to form a load conduction model of the mooring equipment, including: Disassemble the overall structure of the mooring equipment to identify the core structural components, which include load-bearing components that are in direct contact with wind and waves, load-transmitting components, and load-bearing support components. Analyze the physical properties of each core structural component, and determine the load receiving properties of each core structural component based on the physical properties. The load receiving properties include the load types that the core structural component can receive, the upper limit of load receiving, and the deformation characteristics after receiving the load. Through structural mechanics analysis, the connection methods between the core structural components are determined. These connection methods include fixed connections, movable connections, and elastic connections, with different connection methods corresponding to different load conduction characteristics. Based on the connection method and load conduction characteristics, the load conduction path between adjacent core structural components is determined. The load conduction path is marked with the conduction direction from the starting core structural component to the receiving core structural component and the load loss during the conduction process. Each load conduction path is labeled with parameters, including the conduction efficiency, conduction delay, and load conversion ratio during conduction. Position each core structural component according to its spatial distribution location, and mark the position coordinates of each core structural component and the relative distance between the core structural components in the spatial coordinate system; Based on the location coordinates of the core structural components, the load conduction path, and the load conduction path parameters, a topology of the load conduction network is constructed. In the topology of the load conduction network, each node represents a core structural component, and the connection between nodes represents the load conduction path. By supplementing the load receiving attributes and load transmission path parameters of each core structural component in the topology of the load transmission network, a load transmission model of the mooring equipment is formed, which is used to present the transmission process of load within the mooring equipment.

2. The method for automatic prediction of mooring load based on wind and wave big data according to claim 1, characterized in that, For each type of wind and wave action, the action parameters under that type of wind and wave action are extracted, including: For a single type of wind and wave action, select the wind and wave action data records of all monitoring points under that single type of wind and wave action. The wind and wave impact data records for each monitoring point are arranged in chronological order of collection time, and the impact intensity value corresponding to each time point is extracted. Calculate the difference in the intensity values ​​of adjacent time nodes. This difference is used to determine the trend of the intensity change of this single wind and wave type at the monitoring point based on its positive or negative change and the magnitude of the change. Extract the action direction records for each time node under the single wind and wave action type at each monitoring point, count the duration of the same action direction, and determine the action direction shift based on the change in duration. The wind and wave action data records for each monitoring point under the single wind and wave action type are divided into the same preset time window, and the number of times the single wind and wave action type occurs within each preset time window is counted. The frequency of the single wind and wave action at the monitoring point is calculated based on the number of occurrences within each preset time window and the duration of the preset time window. The frequency of action at different monitoring points is arranged according to the spatial distribution of the monitoring points to form a frequency distribution map of the single wind and wave action type. By integrating the intensity change trend, direction shift, and frequency distribution of each monitoring point, the action parameters corresponding to this single type of wind and wave action are formed.

3. The method for automatic prediction of mooring load based on wind and wave big data according to claim 2, characterized in that, The determination of load conduction paths between adjacent core structural components based on connection methods and load conduction characteristics includes: For a pair of adjacent core structural components, the core structural component that bears the direct impact of wind and waves is identified as the starting core structural component, and the core structural component that receives the load transmitted by the starting core structural component is identified as the receiving core structural component. The load conduction characteristics corresponding to the connection methods of the starting core structural component and the receiving core structural component are analyzed. The load conduction characteristics of the fixed connection are that the load is transmitted without directional offset. The load conduction characteristics of the movable connection are that the load can be transmitted along the connection axis. The load conduction characteristics of the elastic connection are that there is elastic buffering during the load transmission process. The specific transmission direction of the load from the starting core structural component to the receiving core structural component is determined based on the load transmission characteristics. The transmission direction must be consistent with the connecting axis direction of the core structural component and the structural force direction of the core structural component. Select the historical load transfer data of the adjacent core structural component, and extract the output load value of the starting core structural component and the input load value of the receiving core structural component from the historical load transfer data. Calculate the difference between the output load value of the initial core structure component and the input load value of the receiving core structure component. This difference represents the load loss value during the conduction process. Statistically analyze the loss values ​​of multiple sets of historical load transfer data, and analyze the correlation between the loss values ​​and the output load values ​​and conduction direction of the initial core structural components; Determine the load loss patterns under different output load values ​​and different conduction directions based on the correlation relationships; The direction of conduction and the law of load loss are marked on the load conduction path of the adjacent core structural components to form a load conduction path description.

4. The method for automatic prediction of mooring load based on wind and wave big data according to claim 1, characterized in that, The process involves inputting the wind-wave action chain into the load conduction network, combining historical wind-wave action chains and historical load conduction data to establish a dynamic response mapping model between wind-wave action and load conduction. This model is used to output the corresponding load conduction distribution results based on the input wind-wave action parameters, including: Collect historical wind and wave action data for the target sea area within a specified time period, and construct historical wind and wave action chains, with each historical wind and wave action chain corresponding to a historical time period. Collect historical load transmission data of the mooring equipment during the aforementioned historical period, including the load values ​​of each core structural component at different times and the load changes along the load transmission path; Each historical wind and wave action chain is associated with the corresponding historical load transmission data to form a historical association dataset, and the load values ​​of each core structural component corresponding to each historical wind and wave action unit are marked. Extract the action parameters of historical wind and wave action chains and the load transmission distribution characteristics from the corresponding historical load transmission data from the historical associated dataset. The load transmission distribution characteristics include the load proportion of each core structural component and the utilization rate of the load transmission path. Construct a mapping relationship training sample set, wherein each sample in the mapping relationship training sample set contains a set of wind and wave action parameters and corresponding load conduction distribution characteristics; A mapping relationship modeling method was selected, and based on time-series correlation analysis, the dynamic correlation between the changes in wind and wave action parameters over time and the changes in load conduction distribution characteristics was captured. The mapping relationship modeling method is trained using a mapping relationship training sample set. The correlation parameters in the mapping relationship modeling process are adjusted to ensure that the output load transmission distribution characteristics are consistent with the actual characteristics in the mapping relationship training sample set. The trained mapping model is combined with the topology of the load transmission network, integrating the load receiving attributes and load transmission path parameters of each core structural component to form a dynamic response mapping model of wind and wave action and load transmission. Input the test wind and wave action chain to the dynamic response mapping model of wind and wave action and load conduction, verify the consistency between the output load conduction distribution results and the actual load conduction data, and optimize the correlation parameters based on the verification results.

5. The method for automatic prediction of mooring load based on wind and wave big data according to claim 4, characterized in that, The step of training the mapping relationship modeling method using a mapping relationship training sample set and adjusting the association parameters in the mapping relationship modeling process includes: The training sample set of the mapping relationship is divided into a training subset and a validation subset. The training subset is used for training the parameters of the mapping relationship model, and the validation subset is used for validating the parameters of the mapping relationship model. The association parameters of the initial mapping relationship model are included, such as the weight coefficient of the wind and wave action parameters, the influence coefficient of the load transmission path, and the attenuation coefficient of the time series association. The wind and wave action parameters of the first sample in the training subset are input into the mapping relationship model, and the corresponding load transmission distribution prediction features are calculated and output based on the initial correlation parameters. Extract the actual load conduction distribution features corresponding to the sample, calculate the degree of difference between the predicted load conduction distribution features and the actual load conduction distribution features, and measure the degree of difference by the sum of the numerical deviations of each dimension of the features; The correlation parameters of the mapping relationship model are adjusted according to the degree of difference. If the load ratio of any core structural component in the load transmission distribution prediction characteristics is higher than that in the actual load transmission distribution characteristics, the weight coefficient of the wind and wave action parameter corresponding to that core structural component is reduced; if the predicted value of the load transmission path utilization rate is lower than the actual value, the influence coefficient of that load transmission path is increased. The next sample in the training subset is input into the adjusted mapping model, and the process of calculating the load transmission distribution prediction features, measuring the degree of difference, and adjusting the correlation parameters is repeated. After traversing all samples in the training subset, the samples in the validation subset are input into the current mapping model to calculate the overall degree of difference on the validation subset. If the overall difference is higher than the preset standard, continue to use the training subset for multiple rounds of parameter adjustment until the overall difference on the validation subset is lower than the preset standard. Record the final adjusted association parameters, fix these parameters into the mapping relationship model, and complete the training process.

6. The method for automatic prediction of mooring load based on wind and wave big data according to claim 1, characterized in that, The process of acquiring real-time wind and wave data of the target sea area, converting the real-time wind and wave data into a real-time wind and wave action chain, inputting the real-time wind and wave action chain into a dynamic response mapping model of wind and wave action and load transmission, obtaining the real-time load transmission distribution result through dynamic correlation calculation, and determining the mooring load prediction value based on the real-time load transmission distribution result includes: Activate the real-time wind and wave monitoring equipment in the target sea area and acquire real-time wind and wave data at the preset acquisition frequency, including the type, intensity, direction and duration of wind and wave action at the current moment. Real-time classification of wind and wave impact data is performed to determine the type of wind and wave impact corresponding to the current real-time wind and wave impact data. Extract the action parameters from the current real-time wind and wave action data, including real-time action intensity changes, real-time action direction shifts, and real-time action frequency distribution; The real-time wind and wave action types and corresponding action parameters collected in chronological order are connected in series to form a real-time wind and wave action unit, and each real-time wind and wave action unit corresponds to one collection cycle. Multiple consecutive real-time wind and wave action units are integrated in chronological order of data acquisition to form a real-time wind and wave action chain; The real-time wind and wave action chain is organized according to the input format required by the dynamic response mapping model of wind and wave action and load transmission. The organized real-time wind and wave action chain is input into the dynamic response mapping model of wind and wave action and load transmission. The internally trained mapping relationship model is called, and dynamic correlation calculation is performed in combination with the topology of the load transmission network and the attributes of the core structural components. During the calculation, based on the parameters of each real-time wind and wave action unit in the real-time wind and wave action chain, the load values ​​of each core structural component and the load changes of each load transmission path are calculated to obtain the real-time load transmission distribution results. Extract the real-time load values ​​of each core structural component from the real-time load transmission and distribution results, and sum the load values ​​of all core structural components to obtain the total load value; Based on the sum of the load values ​​and the load proportion of each core structural component, the overall mooring load prediction value of the mooring equipment is determined.

7. The method for automatic prediction of mooring load based on wind and wave big data according to claim 6, characterized in that, During the calculation process, based on the parameters of each real-time wind and wave action unit in the real-time wind and wave action chain, the load values ​​of each core structural component and the load changes of each load transmission path are calculated to obtain the real-time load transmission distribution results, including: Extract the first real-time wind and wave action unit from the real-time wind and wave action chain, and obtain the wind and wave action type and action parameters corresponding to the first real-time wind and wave action unit; The internal mapping relationship model is invoked, and the parameters of the first real-time wind and wave action unit are input into the mapping relationship model. Based on the trained correlation parameters, the initial load values ​​of each core structural component under the action of the first real-time wind and wave action unit are calculated. Based on the topology of the load conduction network, determine the load conduction path between each core structural component, and calculate the initial load conduction value from the starting core structural component to the receiving core structural component according to the conduction direction marked on the load conduction path and the load loss law. Record the load conduction value changes on each load conduction path, including the output load of the starting core structural component, the loss load in the load conduction path, and the input load of the receiving core structural component; After completing the load calculation for the first real-time wind and wave action unit, extract the next real-time wind and wave action unit in the real-time wind and wave action chain and input the parameters of the next real-time wind and wave action unit into the mapping relationship model. Based on the final load values ​​of each core structural component under the action of the previous real-time wind and wave action unit, calculate the new load values ​​of each core structural component under the action of the current real-time wind and wave action unit. The newly added load value is superimposed with the final load value of the previous moment to obtain the real-time load value of each core structural component under the action of the current real-time wind and wave action unit. Repeat the above steps to complete the load calculation of all real-time wind and wave action units in the real-time wind and wave action chain in sequence, and obtain the load values ​​of each core structural component and the load changes of each load transmission path corresponding to each real-time wind and wave action unit. The load calculation results corresponding to all real-time wind and wave action units are integrated in chronological order to form a real-time load transmission distribution result that includes the load values ​​of each core structural component at different times and the load changes of each load transmission path at different times.

8. An automatic mooring load prediction system based on wind and wave big data, characterized in that, include: processor; A machine-readable storage medium for storing machine-executable instructions of the processor; The processor is configured to execute the automatic mooring load prediction method based on big data of wind and waves as described in any one of claims 1 to 7 by executing the machine-executable instructions.

Images