Method and System for Analyzing the Service Life of a Floating Platform Based on Dynamic Loads of Wind and Waves

Abstract

The present invention provides a method and system for analyzing the lifespan of a floating platform based on dynamic loads of wind, waves, and currents. The method includes: obtaining in real time the monitoring data of the wind field, wave field, and current field in the ocean area where the floating platform to be analyzed is located, and extracting the initial spatio-temporal sequence; extracting coupling features; generating grids with different degrees of fineness according to the resolution at different positions to obtain a multi-resolution feature distribution; analyzing the time sequence of the feature distribution to obtain a load spectrum; using finite element analysis to calculate the stress distribution of multiple key nodes of the platform to obtain the stress state data of the platform; identifying the fatigue cycle times of the key nodes according to the stress state data; and combining with a preset material fatigue curve to obtain the remaining lifespan of the key nodes, and further obtaining the fatigue lifespan of the floating platform to be analyzed. The present invention takes into account the dynamic coupling load action of wind-wave-current, reduces the deviation between the lifespan analysis result and the actual service state, and improves the accuracy of the analysis result.

Description

Technical Field

[0001] The present invention relates to the field of marine data processing, and particularly to a method and system for analyzing the lifespan of a floating platform based on dynamic loads of wind, wave, and current. Background Art

[0002] The field of ocean engineering plays a crucial role in global energy development and resource utilization. A floating platform is the core equipment for deep-sea energy development, and its design and operation and maintenance directly affect work safety and economic benefits. During the long-term service of a floating platform, it often needs to face a complex marine environment, and the coupled action of factors such as wind, wave, and current poses extremely high requirements for the fatigue and lifespan prediction of the platform structure.

[0003] However, existing methods often rely on simplified models (such as a wave field model depending on a single factor) or low-resolution data analysis when dealing with the coupled loads of wind, wave, and current, and it is difficult to accurately capture the spatio-temporal characteristics of the dynamic interaction between two or more factors of wind, wave, and current. This leads to a deviation between load prediction and the actual service state, thereby affecting the accuracy of platform lifespan analysis. Summary of the Invention

[0004] The present invention provides a method and system for analyzing the lifespan of a floating platform based on dynamic loads of wind, wave, and current to solve the technical problem of how to improve the accuracy of floating platform lifespan analysis.

[0005] To solve the above technical problem, an embodiment of the present invention provides a method for analyzing the lifespan of a floating platform based on dynamic loads of wind, wave, and current, including:

[0006] Real-time obtaining monitoring data of the wind field, wave field, and current field in the ocean area where the floating platform to be analyzed is located, and extracting an initial spatio-temporal sequence from the monitoring data; wherein, the floating platform to be analyzed includes a plurality of key nodes;

[0007] Extracting coupling features from the initial spatio-temporal sequence; based on the coupling features, according to the resolution of the monitoring data at different positions in the ocean area where the floating platform to be analyzed is located, generating grids with different levels of fineness respectively, and obtaining a multi-resolution feature distribution according to the grid generation result; wherein, the grid fineness corresponding to the area with a relatively large resolution is greater than the grid fineness corresponding to the area with a relatively small resolution;

[0008] Analyzing the time sequence of the multi-resolution feature distribution to obtain a load spectrum;

[0009] Based on the load spectrum, using finite element analysis to calculate the stress distribution of the plurality of key nodes to obtain stress state data of the floating platform to be analyzed;

[0010] Based on the stress state data, the fatigue cycle times of the key nodes are identified; and in combination with the preset material fatigue curve, the remaining life of the key nodes is obtained, and then based on each of the remaining lives, the fatigue life of the floating platform to be analyzed is obtained.

[0011] As a preferred solution, the fineness is set according to the element length, the number of grids, and the grid density;

[0012] Based on the coupling characteristics, according to the resolution of the monitoring data at different positions in the ocean area where the floating platform to be analyzed is located, grids with different fineness levels are respectively generated, and a multi-resolution feature distribution is obtained according to the grid generation results, including:

[0013] The ocean area where the floating platform to be analyzed is located is divided into multiple sub-areas;

[0014] Multiple calculation queues are constructed, and each calculation queue corresponds to a sub-area respectively;

[0015] According to the resolution of the monitoring data of each sub-area, the element length, the number of grids, and the grid density of each sub-area are determined;

[0016] Control the multiple calculation queues to respectively generate the target grids of each sub-area according to the element length, the number of grids, and the grid density of each sub-area;

[0017] Based on the coupling characteristics and the target grids of each sub-area, the multi-resolution feature distribution is obtained.

[0018] As a preferred solution, the step of controlling the multiple calculation queues to respectively generate the target grids of each sub-area according to the element length, the number of grids, and the grid density of each sub-area includes:

[0019] Control the multiple calculation queues to respectively and preliminarily generate grid surface information according to the element length, the number of grids, and the grid density of each sub-area;

[0020] Control the multiple calculation queues to generate three-dimensional grids of each sub-area based on the grid surface information and obtain grid body information;

[0021] According to the grid body information, the vertices and volume elements of each three-dimensional grid are obtained; and the vertices and volume elements of each three-dimensional grid are numbered to obtain the unique strings of each vertex and the unique strings of each volume element;

[0022] Control synchronous communication between the multiple calculation queues so that the volume elements and unique strings between adjacent sub-areas are projected onto each other to obtain projection information;

[0023] Optimize the three-dimensional grids of each sub-area according to the projection information to obtain the target grids of each sub-area.

[0024] As a preferred solution, extracting the initial spatio-temporal sequence from the monitoring data includes:

[0025] Performing normalization processing on the monitoring data of the wind field, wave field, and current field to obtain normalized data;

[0026] Using a preset convolutional neural network to extract wind field features, wave field features, and current field features from the normalized data to obtain the initial spatio-temporal sequence.

[0027] As a preferred solution, extracting the coupling features from the initial spatio-temporal sequence includes:

[0028] Extracting a preliminary feature set of the coupling of the wind field, wave field, and current field from the initial spatio-temporal sequence;

[0029] When the feature dimension of the preliminary feature set is greater than a preset dimension threshold, performing dimensionality reduction processing on the preliminary feature set to obtain dimensionality-reduced data;

[0030] Analyzing the time series of the dimensionality-reduced data and enhancing the correlation features regarding the dynamic characteristics of the interaction of the wind field, wave field, and current field to obtain the coupling features.

[0031] As a preferred solution, the floating platform to be analyzed is connected to at least one anchor chain; based on the load spectrum, using finite element analysis to calculate the stress distribution of the multiple key nodes to obtain the stress state data of the floating platform to be analyzed, including:

[0032] Constructing a platform model of the floating platform to be analyzed; respectively constructing anchor chain models of each anchor chain, and embedding all the anchor chain models into the platform model to obtain a target model;

[0033] Constructing an external excitation based on the load spectrum; according to the target model and the external excitation, combining with preset boundary conditions for finite element analysis to obtain the stress distribution of each key node;

[0034] Analyzing the time variation trend and spatial variation trend of the stress of each key node according to the stress distribution of each key node;

[0035] Based on the time variation trend and spatial variation trend of the stress of each key node, obtaining a dynamic stress distribution map of the floating platform to be analyzed, and obtaining the stress state data according to the dynamic stress distribution map.

[0036] As a preferred solution, the floating platform life analysis method further includes:

[0037] Responding to an analysis instruction of a target node, where the analysis instruction includes a calculation accuracy requirement;

[0038] According to the analysis instruction, extract the data sequence corresponding to the position of the target node from the stress state data;

[0039] And according to the calculation accuracy requirement, analyze the data sequence to obtain the stress analysis result of the target node.

[0040] As a preferred solution, based on the stress state data, identify the fatigue cycle times of the key nodes; and combine with the preset material fatigue curve to obtain the remaining life of the key nodes, and then based on each of the remaining lives, obtain the fatigue life of the floating platform to be analyzed, including:

[0041] Divide the floating platform to be analyzed into multiple sub-units, and obtain the calculation accuracy of each sub-unit;

[0042] According to the calculation accuracy of each sub-unit, extract the stress data sequences of each key node from the stress state data;

[0043] Eliminate the noise of the stress data sequence through a low-pass filter to obtain denoised data; and perform moving average processing on the denoised data to obtain smoothed data;

[0044] According to the preset stress amplitude and stress mean value, perform statistics on the smoothed data to obtain the fatigue cycle times of each key node;

[0045] According to the fatigue cycle times and the material fatigue curve, combine with the preset single damage value, and use Miner's linear cumulative damage theory to calculate the cumulative damage of the key nodes;

[0046] Obtain the preset initial life of the key node, subtract the cumulative damage from the initial life, and calculate the remaining life of the key node; based on each of the remaining lives, obtain the fatigue life of the floating platform to be analyzed.

[0047] Correspondingly, the present invention application also provides a floating platform life analysis system based on dynamic wind and wave loads, including a spatio-temporal sequence extraction module, a feature distribution generation module, a load spectrum analysis module, a stress acquisition module, and a life analysis module; wherein,

[0048] The spatio-temporal sequence extraction module is used to obtain the monitoring data of the wind field, wave field, and flow field in the ocean area where the floating platform to be analyzed is located in real time, and extract the initial spatio-temporal sequence from the monitoring data; wherein, the floating platform to be analyzed includes a plurality of key nodes;

[0049] The feature distribution generation module is used to extract coupled features from the initial spatio-temporal sequence; based on the coupled features, according to the resolution of the monitoring data at different positions in the ocean area where the floating platform to be analyzed is located, grids with different degrees of fineness are respectively generated, and a multi-resolution feature distribution is obtained according to the grid generation result; wherein, the grid fineness corresponding to the area with a relatively large resolution is greater than the grid fineness corresponding to the area with a relatively small resolution.

[0050] The load spectrum analysis module is used to analyze the time series of the multi-resolution feature distribution to obtain a load spectrum.

[0051] The stress acquisition module is used to calculate the stress distribution of the multiple key nodes based on the load spectrum by using finite element analysis, and obtain the stress state data of the floating platform to be analyzed.

[0052] The life analysis module is used to identify the fatigue cycle times of the key nodes according to the stress state data; and in combination with a preset material fatigue curve, obtain the remaining life of the key nodes, and further obtain the fatigue life of the floating platform to be analyzed based on each of the remaining lives.

[0053] As a preferred solution, the degree of fineness is set according to the element length, the number of grids, and the grid density.

[0054] The feature distribution generation module, based on the coupled features, according to the resolution of the monitoring data at different positions in the ocean area where the floating platform to be analyzed is located, respectively generates grids with different degrees of fineness, and obtains a multi-resolution feature distribution according to the grid generation result, including:

[0055] The feature distribution generation module divides the ocean area where the floating platform to be analyzed is located into multiple sub-areas.

[0056] Construct multiple calculation queues, and each calculation queue corresponds to a sub-area respectively.

[0057] According to the resolution of the monitoring data of each sub-area, determine the element length, the number of grids, and the grid density of each sub-area.

[0058] Control the multiple calculation queues to respectively generate the target grids of each sub-area according to the element length, the number of grids, and the grid density of each sub-area.

[0059] Based on the coupled features and the target grids of each sub-area, obtain the multi-resolution feature distribution.

[0060] As a preferred solution, the feature distribution generation module controls the multiple calculation queues to respectively generate the target grids of each sub-area according to the element length, the number of grids, and the grid density of each sub-area, including:

[0061] The feature distribution generation module controls the multiple computing queues to initially generate grid surface information according to the unit length, the number of grids, and the grid density of each sub-region respectively;

[0062] controls the multiple computing queues to generate three-dimensional grids for each sub-region based on the grid surface information and obtain grid body information;

[0063] According to the grid body information, obtain the vertices and volume elements of each three-dimensional grid; and number the vertices and volume elements of each three-dimensional grid to obtain the unique strings of each vertex and the unique strings of each volume element;

[0064] controls synchronous communication among the multiple computing queues so that the volume elements and unique strings between adjacent sub-regions are projected onto each other to obtain projection information;

[0065] Optimize the three-dimensional grids of each sub-region according to the projection information to obtain the target grids of each sub-region.

[0066] As a preferred solution, the spatio-temporal sequence extraction module extracts an initial spatio-temporal sequence from the monitoring data, including:

[0067] The spatio-temporal sequence extraction module performs standardization processing on the monitoring data of the wind field, wave field, and flow field to obtain standardized data;

[0068] Uses a preset convolutional neural network to extract wind field features, wave field features, and flow field features from the standardized data to obtain the initial spatio-temporal sequence.

[0069] As a preferred solution, the feature distribution generation module extracts coupling features from the initial spatio-temporal sequence, including:

[0070] The feature distribution generation module extracts a preliminary feature set of the coupling of the wind field, wave field, and flow field from the initial spatio-temporal sequence;

[0071] When the feature dimension of the preliminary feature set is greater than a preset dimension threshold, perform dimensionality reduction processing on the preliminary feature set to obtain dimensionality-reduced data;

[0072] Analyze the time series of the dimensionality-reduced data and enhance the correlation features regarding the dynamic characteristics of the interaction of the wind field, wave field, and flow field to obtain the coupling features.

[0073] As a preferred solution, the floating platform to be analyzed is connected to at least one anchor chain; the stress acquisition module calculates the stress distribution of the multiple key nodes by using finite element analysis based on the load spectrum to obtain the stress state data of the floating platform to be analyzed, including:

[0074] The stress acquisition module constructs a platform model of the floating platform to be analyzed; constructs chain models of each of the anchor chains respectively, and embeds all the chain models into the platform model to obtain a target model;

[0075] Constructs an external excitation based on the load spectrum; performs finite element analysis according to the target model and the external excitation in combination with preset boundary conditions to obtain the stress distribution of each key node;

[0076] Analyzes the time variation trend and spatial variation trend of the stress of each key node according to the stress distribution of each key node;

[0077] Obtains a dynamic stress distribution map of the floating platform to be analyzed based on the time variation trend and spatial variation trend of the stress of each key node, and obtains the stress state data according to the dynamic stress distribution map.

[0078] As a preferred solution, the floating platform life analysis system further includes a target node analysis module, and the target node analysis module is used for:

[0079] Responds to an analysis instruction of a target node, and the analysis instruction includes a calculation accuracy requirement;

[0080] Extracts a data sequence corresponding to the position of the target node from the stress state data according to the analysis instruction;

[0081] And analyzes and obtains a stress analysis result of the target node based on the data sequence according to the calculation accuracy requirement.

[0082] As a preferred solution, the life analysis module identifies the fatigue cycle times of key nodes according to the stress state data; combines with a preset material fatigue curve to obtain the remaining life of the key nodes, and further obtains the fatigue life of the floating platform to be analyzed based on each of the remaining lives, including:

[0083] The life analysis module divides the floating platform to be analyzed into multiple sub-units, and obtains the calculation accuracy of each sub-unit;

[0084] Extracts the stress data sequences of each key node from the stress state data according to the calculation accuracy of each sub-unit;

[0085] Eliminates the noise of the stress data sequence through a low-pass filter to obtain denoised data; and performs a moving average process on the denoised data to obtain smoothed data;

[0086] Counts the smoothed data according to a preset stress amplitude and stress mean value to obtain the fatigue cycle times of each key node;

[0087] According to the number of fatigue cycles and the material fatigue curve, combined with a preset single damage value, the Miner linear cumulative damage theory is used to calculate the cumulative damage of key nodes;

[0088] Obtain the preset initial life of the key node, subtract the cumulative damage from the initial life, and calculate the remaining life of the key node; based on each of the remaining lives, obtain the fatigue life of the floating platform to be analyzed.

[0089] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0090] In the present invention application, by obtaining the monitoring data of the wind field, wave field and current field in the ocean area where the floating platform to be analyzed is located, extracting the initial spatio-temporal sequence, and obtaining the coupling characteristics, the dynamic characteristics under the interaction of multiple factors of the wind field, wave field and current field can be determined; further, according to the resolution of the monitoring data at different positions in the ocean area where the floating platform to be analyzed is located, grids with different levels of fineness are respectively generated. Relatively fine grids are generated at positions with relatively large resolution, and relatively sparse grids are generated at positions with relatively small resolution, obtaining a multi-resolution feature distribution, which can realize the targeted distribution of the resolution of different regions, realize the dynamic adaptation of the ocean area where the floating platform to be analyzed is located under the action of wind-wave-current coupling loads, and then obtain an accurate load spectrum, and on this basis, an effective balance between simulation accuracy and calculation cost is achieved; in addition, since the load spectrum is obtained through time series analysis of the multi-resolution feature distribution, the time evolution characteristics of the dynamic coupling loads of the wind field, wave field and current field can be extracted. When calculating the stress distribution of the key nodes of the floating platform to be analyzed, accurate stress distribution data can be obtained, and then the remaining life of the key nodes and the fatigue life of the floating platform can be predicted. Compared with the analysis of a single factor in the prior art, the present invention application considers the action of the dynamic coupling loads of wind-wave-current, reduces the deviation between the analysis result of the floating platform life and the actual service state, and improves the accuracy of the analysis result; finally, in this application, the stress distributions of multiple key nodes are calculated to obtain the stress state data of the floating platform to be analyzed as a whole. By using the stress state data of the floating platform with integrity and smoothness to identify the number of fatigue cycles of the key nodes and then obtain the remaining life, compared with the method of directly predicting the remaining life of the key nodes using the stress distribution of the key nodes, it can avoid the excessive deviation of the remaining life identified by each key node, resulting in the distortion of the fatigue life data of the floating platform finally analyzed. That is, it is equivalent to using the stress state data of the floating platform to be analyzed as a whole for similar smoothing processing before identifying the remaining life of each key node, controlling the remaining life of each key node within a relatively smaller interval, so as to analyze and obtain a more accurate and stable fatigue life of the floating platform to be analyzed. Description of the Drawings

[0091] Figure 1 : Schematic flow diagram of an embodiment of the floating platform life analysis method based on wind-wave-flow dynamic loads provided by this application of the present invention.

[0092] Figure 2 : Schematic flow diagram of another embodiment of the floating platform life analysis method based on wind-wave-flow dynamic loads provided by this application of the present invention.

[0093] Figure 3 : Schematic structural diagram of an embodiment of the floating platform life analysis system based on wind-wave-flow dynamic loads provided by this application of the present invention. Detailed implementation manners

[0094] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0095] Embodiment 1:

[0096] Please refer to Figure 1 , Figure 1 A floating platform life analysis method based on wind-wave-flow dynamic loads provided for an embodiment of the present invention, including steps S101 to S105; wherein,

[0097] Step S101, obtain the monitoring data of the wind field, wave field, and flow field in the ocean area where the floating platform to be analyzed is located in real time, and extract the initial spatio-temporal sequence from the monitoring data.

[0098] Among them, the floating platform to be analyzed includes a plurality of key nodes, and the key nodes can be pre-configured or selected.

[0099] In this step, the floating platform to be analyzed is located in an ocean area, and the floating platform to be analyzed can be connected to at least one anchor chain.

[0100] This embodiment can obtain the monitoring data of the wind field, wave field, and flow field in the ocean area where the floating platform to be analyzed is located in real time through a preset sensor network.

[0101] The sensor network includes various different types of sensors. Therefore, for the monitoring data collected by the sensor network, a multi-source data fusion algorithm can be used to preprocess it, and then the initial spatio-temporal sequence can be extracted.

[0102] In a preferred implementation manner, the extracting the initial spatio-temporal sequence from the monitoring data includes:

[0103] Standardize the monitoring data of the wind field, wave field, and current field so that the monitoring data of each field can be converted into a unified format to obtain standardized data;

[0104] Use a preset convolutional neural network to extract wind field features, wave field features, and current field features from the standardized data to obtain the initial spatio-temporal sequence.

[0105] Exemplarily, this convolutional neural network can use a three-layer two-dimensional convolutional neural network architecture. The first layer uses 32 5×5 convolutional kernels with ReLU activation functions to extract local spatial features. The second layer captures multi-field coupling features through 64 3×3 convolutional kernels. The third layer uses 128 1×1 convolutional kernels for feature compression, and each layer is followed by a 2×2 max pooling layer to reduce the dimension, thereby extracting wind field features, wave field features, and current field features, and then obtaining the initial spatio-temporal sequence.

[0106] Step S102, extract coupling features from the initial spatio-temporal sequence; based on the coupling features, according to the resolution of the monitoring data at different positions in the ocean area where the floating platform to be analyzed is located, generate grids with different levels of fineness respectively, and obtain a multi-resolution feature distribution according to the grid generation results; wherein, the grids corresponding to the areas with relatively large resolution have a greater level of fineness than the grids corresponding to the areas with relatively small resolution.

[0107] In a preferred embodiment, the level of fineness is set according to parameters such as cell length, number of grids, and grid density.

[0108] The extraction of coupling features from the initial spatio-temporal sequence includes:

[0109] Extract a preliminary feature set of the coupling of the wind field, wave field, and current field from the initial spatio-temporal sequence;

[0110] When the feature dimension of the preliminary feature set is greater than a preset dimension threshold, perform dimensionality reduction processing on the preliminary feature set to obtain dimensionality-reduced data (when the feature dimension of the preliminary feature set is greater than the preset dimension threshold, no dimensionality reduction operation is required, and subsequent analysis can be directly performed);

[0111] Analyze the time series of the dimensionality-reduced data, and enhance the correlation features regarding the dynamic characteristics of the interaction of the wind field, wave field, and current field (for example, highlight the correlation features or expand the entire dataset) to obtain the coupling features.

[0112] The purpose of the operation of enhancing the correlation features in this embodiment is to highlight the coupling effect of the wind field, wave field, and current field. Exemplarily, in some application instances, the coupling features of the present invention application can be further analyzed, such as performing clustering and other operations to obtain classical patterns of the coupling of the wind field, wave field, and current field, etc.

[0113] Further, please refer to Figure 2 , and based on the embodiment shown in Figure 1 , a flow chart of step S101 of another embodiment of the method for analyzing the service life of a floating platform based on dynamic loads of wind and waves is provided. It should be noted that the steps identical to Figure 2 are not described herein again. Figure 1

[0114] As shown in Figure 2 , in step S101, based on the coupling characteristics, according to the resolution of the monitoring data at different positions in the ocean area where the floating platform to be analyzed is located, grids with different degrees of refinement are respectively generated, and a multi-resolution feature distribution is obtained according to the grid generation result, including steps S201 to S205. Each step is described in detail as follows:

[0115] Step S201: Divide the ocean area where the floating platform to be analyzed is located into multiple sub-areas.

[0116] Step S202: Construct multiple computing queues, each computing queue corresponding to a sub-area, and the tasks of each computing queue can be executed by different servers or different processors or different cores.

[0117] Step S203: Determine the unit length, the number of grids, and the grid density of each sub-area according to the resolution of the monitoring data of each sub-area, that is, determine the refinement degree of generating the grids corresponding to the sub-areas.

[0118] Step S204: Control the multiple computing queues to respectively generate the target grids of each sub-area according to the unit length, the number of grids, and the grid density of each sub-area.

[0119] Step S205: Based on the coupling characteristics and the target grids of each sub-area, obtain the multi-resolution feature distribution.

[0120] In this embodiment, grids with different refinement degrees are adaptively generated according to the resolution of the monitoring data of each sub-area. In this way, for sub-areas with high resolution, relatively fine grids can be generated, and for sub-areas with low resolution, relatively sparse grids can be generated, obtaining a multi-resolution feature distribution, which can realize the targeted distribution of the resolution of different areas and achieve the dynamic adaptation of the ocean area where the floating platform to be analyzed is located under the action of wind-wave-current coupling loads (for example, high-resolution data can simulate fine loads, and low-resolution data does not need to generate loads with such high precision, but can save computing resources by reducing the grid refinement degree), thereby obtaining an accurate load spectrum as a whole, and achieving an effective balance between simulation accuracy and computing cost on this basis.

[0121] Further, controlling the multiple computing queues to generate the target grids for each sub-region according to the unit length, the number of grids, and the grid density of each sub-region respectively includes:

[0122] Controlling the multiple computing queues to respectively generate grid surface information according to the unit length, the number of grids, and the grid density of each sub-region;

[0123] Controlling the multiple computing queues to generate three-dimensional grids for each sub-region based on the grid surface information and obtain grid body information;

[0124] According to the grid body information, obtaining the vertices and volume elements of each three-dimensional grid; and numbering the vertices and volume elements of each three-dimensional grid to obtain the unique strings of each vertex and the unique strings of each volume element;

[0125] Controlling synchronous communication among the multiple computing queues so that the volume elements and unique strings between adjacent sub-regions are projected onto each other to obtain projection information;

[0126] Optimizing the three-dimensional grids of each sub-region according to the projection information to obtain the target grids of each sub-region.

[0127] In this embodiment, by controlling synchronous communication among the multiple computing queues so that the volume elements and unique strings between adjacent sub-regions are projected onto each other to obtain projection information, and using the projection information to optimize the three-dimensional grids of each sub-region, compared with the existing technical solution of directly merging each sub-region, the problem of inconsistent interface surfaces between adjacent regions during merging is avoided, and no post-processing steps are required; in addition, the present invention application does not require a merging operation, but uses the projection information to optimize the three-dimensional grids of each sub-region, and then obtains the required multi-resolution feature distribution according to the coupling characteristics and the target grids of each sub-region in subsequent steps, which can avoid a large amount of CPU resources required for the merging operation and has a lower requirement for the memory capacity (the existing technical solution requires a larger memory to accommodate the overall grid after merging).

[0128] Step S103, analyzing the time series of the multi-resolution feature distribution to obtain a load spectrum.

[0129] In a preferred embodiment, when analyzing the time series of the multi-resolution feature distribution, methods such as kernel density estimation can be used to calculate the distribution density in the time dimension, and then generate a load spectrum of the coupling of the wind field, the wave field, and the flow field.

[0130] Further, the frequency-domain characteristics and time-frequency characteristics of the load spectrum can be analyzed, and the uncertainty of the load spectrum can be quantified to determine whether the generated load spectrum meets the requirements, and if not, the load spectrum is regenerated.

[0131] Step S104: Based on the load spectrum, use finite element analysis to calculate the stress distribution of the multiple key nodes, and obtain the stress state data of the floating platform to be analyzed.

[0132] In a preferred embodiment, the step of using finite element analysis to calculate the stress distribution of the multiple key nodes based on the load spectrum and obtaining the stress state data of the floating platform to be analyzed includes:

[0133] Construct the platform model of the floating platform to be analyzed; respectively construct the mooring chain models of each mooring chain, and embed all the mooring chain models into the platform model to obtain the target model;

[0134] Construct an external excitation based on the load spectrum; perform finite element analysis according to the target model and the external excitation in combination with preset boundary conditions to obtain the stress distribution of each key node;

[0135] Analyze the time-varying trend and space-varying trend of the stress of each key node according to the stress distribution of each key node;

[0136] Based on the time-varying trend and space-varying trend of the stress of each key node, obtain the dynamic stress distribution map of the floating platform to be analyzed, and obtain the stress state data according to the dynamic stress distribution map.

[0137] In this embodiment, finite element analysis tools such as Ansys can be used to construct the model. By constructing models for the floating platform to be analyzed and the mooring chains respectively, that is, considering the mutual independence of the floating platform to be analyzed and the mooring chains in the actual marine environment stress during the simulation of the stress distribution, the accuracy of the stress simulation is higher compared with constructing the model as a whole; further, by analyzing the time-varying trend and space-varying trend of the stress of each key node, the dynamic stress distribution map of the floating platform to be analyzed can combine the time-varying characteristics and space-varying characteristics of the stress of each key node, and reflect the stress characteristics of the platform in real time, with higher accuracy.

[0138] Step S105: Identify the fatigue cycle times of the key nodes according to the stress state data; and combine the preset material fatigue curve to obtain the remaining life of the key nodes, and then based on each remaining life, obtain the fatigue life of the floating platform to be analyzed.

[0139] It should be noted that in this application, the stress distribution of multiple key nodes is calculated to obtain the stress state data of the floating platform to be analyzed as a whole. By using the stress state data that has integrity and smoothness of the floating platform to be analyzed, the fatigue cycle times of the key nodes are identified and then the remaining life is obtained. In this way, compared with the method of directly predicting the remaining life of the key nodes using the stress distribution of the key nodes, it can avoid the excessive deviation of the remaining life identified for each key node, resulting in the distortion of the fatigue life data of the floating platform finally analyzed. That is, it is equivalent to performing a similar smoothing process using the stress state data of the floating platform to be analyzed as a whole before identifying the remaining life of each key node, controlling the remaining life of each key node within a relatively smaller range, so as to analyze and obtain a more accurate and stable fatigue life of the floating platform to be analyzed.

[0140] In a preferred implementation manner, as described above, since this application calculates the stress distribution of multiple key nodes to obtain the stress state data of the floating platform to be analyzed as a whole, and uses the stress state data that has integrity and smoothness of the floating platform to be analyzed to identify the fatigue cycle times of the key nodes and then obtain the remaining life. When obtaining the stress state data of the floating platform to be analyzed as a whole according to the stress distribution of multiple key nodes, it is not a simple splicing between nodes and has a certain smoothness. Therefore, for the target nodes that are not key nodes, the floating platform life analysis method further includes:

[0141] Respond to the analysis instruction of the target node, and the analysis instruction includes the calculation accuracy requirement;

[0142] According to the analysis instruction, extract the data sequence corresponding to the position of the target node from the stress state data;

[0143] And according to the calculation accuracy requirement, analyze and obtain the stress analysis result of the target node based on the data sequence.

[0144] In a preferred implementation manner, the method of identifying the fatigue cycle times of the key nodes according to the stress state data; and combining with the preset material fatigue curve to obtain the remaining life of the key nodes, and then based on each of the remaining lives, obtaining the fatigue life of the floating platform to be analyzed includes:

[0145] Divide the floating platform to be analyzed into multiple sub-units, and obtain the calculation accuracy of each sub-unit;

[0146] According to the calculation accuracy of each sub-unit, extract the stress data sequences of each key node from the stress state data;

[0147] Eliminate the high-frequency noise of the stress data sequence through a low-pass filter (such as a Butterworth low-pass filter with a cut-off frequency of 5 Hz and an order of 4) to obtain denoised data; and perform a moving average process on the denoised data to obtain smoothed data;

[0148] According to the preset stress amplitude (such as 0.1 Mpa) and stress mean value (such as 0.5 Mpa), perform statistics on the smoothed data to obtain the fatigue cycle times of each key node (for example, identify that there are 128 cycles with an amplitude of 15.3 MPa / mean value of -2.1 MPa and 256 cycles with an amplitude of 8.7 MPa / mean value of 5.4 MPa);

[0149] According to the fatigue cycle times and the material fatigue curve, combined with the preset single damage value, calculate the cumulative damage of the key node using the Miner linear cumulative damage theory;

[0150] Obtain the preset initial life of the key node, subtract the cumulative damage from the initial life, and calculate the remaining life of the key node; based on each remaining life, obtain the fatigue life of the floating platform to be analyzed.

[0151] Correspondingly, as Figure 3 shown, the present invention application also provides a floating platform life analysis system 300 based on wind and wave dynamic loads, including a spatio-temporal sequence extraction module 301, a feature distribution generation module 302, a load spectrum analysis module 303, a stress acquisition module 304, and a life analysis module 305; wherein,

[0152] The spatio-temporal sequence extraction module 301 is used to obtain the monitoring data of the wind field, wave field, and current field in the ocean area where the floating platform to be analyzed is located in real time, and extract the initial spatio-temporal sequence from the monitoring data; wherein, the floating platform to be analyzed includes a plurality of key nodes;

[0153] The feature distribution generation module 302 is used to extract coupling features from the initial spatio-temporal sequence; based on the coupling features, according to the resolution of the monitoring data at different positions in the ocean area where the floating platform to be analyzed is located, generate grids with different degrees of fineness respectively, and obtain a multi-resolution feature distribution according to the grid generation result; wherein, the grid fineness corresponding to the area with a relatively large resolution is greater than the grid fineness corresponding to the area with a relatively small resolution;

[0154] The load spectrum analysis module 303 is used to analyze the time sequence of the multi-resolution feature distribution to obtain a load spectrum;

[0155] The stress acquisition module 304 is used to calculate the stress distribution of the plurality of key nodes based on the load spectrum by finite element analysis to obtain the stress state data of the floating platform to be analyzed;

[0156] The service life analysis module 305 is configured to identify the fatigue cycle times of key nodes according to stress state data, and obtain the remaining service life of key nodes in combination with a preset material fatigue curve. Furthermore, based on the remaining service lives, the fatigue life of the floating platform to be analyzed is obtained.

[0157] As a preferred solution, the fineness is set according to the element length, the number of meshes, and the mesh density.

[0158] Based on the coupled features, the feature distribution generation module 302 generates meshes with different fineness levels respectively according to the resolution of the monitoring data at different positions in the ocean area where the floating platform to be analyzed is located, and obtains a multi-resolution feature distribution according to the mesh generation result, including:

[0159] The feature distribution generation module 302 divides the ocean area where the floating platform to be analyzed is located into multiple sub-areas.

[0160] Construct multiple calculation queues, each calculation queue corresponding to a sub-area respectively.

[0161] According to the resolution of the monitoring data of each sub-area, determine the element length, the number of meshes, and the mesh density of each sub-area.

[0162] Control the multiple calculation queues to generate the target meshes of each sub-area respectively according to the element length, the number of meshes, and the mesh density of each sub-area.

[0163] Based on the coupled features and the target meshes of each sub-area, obtain the multi-resolution feature distribution.

[0164] As a preferred solution, the feature distribution generation module 302 controls the multiple calculation queues to generate the target meshes of each sub-area respectively according to the element length, the number of meshes, and the mesh density of each sub-area, including:

[0165] The feature distribution generation module 302 controls the multiple calculation queues to respectively generate mesh surface information according to the element length, the number of meshes, and the mesh density of each sub-area.

[0166] Control the multiple calculation queues to generate three-dimensional meshes of each sub-area based on the mesh surface information and obtain mesh volume information.

[0167] According to the mesh volume information, obtain the vertices and volume elements of each three-dimensional mesh; and number the vertices and volume elements of each three-dimensional mesh to obtain the unique strings of each vertex and the unique strings of each volume element.

[0168] Control the synchronous communication between the multiple computing queues to project the physical units and uniqueness strings between adjacent sub-regions onto each other, and obtain projection information;

[0169] Optimize the three-dimensional grids of each sub-region according to the projection information to obtain the target grids of each sub-region.

[0170] As a preferred solution, the spatio-temporal sequence extraction module 301 extracts an initial spatio-temporal sequence from the monitoring data, including:

[0171] The spatio-temporal sequence extraction module 301 performs standardization processing on the monitoring data of the wind field, wave field, and flow field to obtain standardized data;

[0172] Use a preset convolutional neural network to extract wind field features, wave field features, and flow field features from the standardized data to obtain the initial spatio-temporal sequence.

[0173] As a preferred solution, the feature distribution generation module 302 extracts coupled features from the initial spatio-temporal sequence, including:

[0174] The feature distribution generation module 302 extracts a preliminary feature set of the coupling of the wind field, wave field, and flow field from the initial spatio-temporal sequence;

[0175] When the feature dimension of the preliminary feature set is greater than a preset dimension threshold, perform dimensionality reduction processing on the preliminary feature set to obtain dimensionality-reduced data;

[0176] Analyze the time series of the dimensionality-reduced data, and perform enhancement processing on the correlation features regarding the dynamic characteristics of the interaction of the wind field, wave field, and flow field to obtain the coupled features.

[0177] As a preferred solution, the floating platform to be analyzed is connected to at least one anchor chain; the stress acquisition module 304 calculates the stress distribution of the multiple key nodes based on the load spectrum by using finite element analysis to obtain the stress state data of the floating platform to be analyzed, including:

[0178] The stress acquisition module 304 constructs a platform model of the floating platform to be analyzed; respectively construct anchor chain models of each anchor chain, and embed all the anchor chain models into the platform model to obtain a target model;

[0179] Construct an external excitation based on the load spectrum; perform finite element analysis according to the target model and the external excitation, combined with preset boundary conditions, to obtain the stress distribution of each key node;

[0180] According to the stress distribution of each key node, analyze the time change trend and space change trend of the stress of each key node;

[0181] Based on the time-varying trend and space-varying trend of the stress at each of the key nodes, a dynamic stress distribution map of the floating platform to be analyzed is obtained, and stress state data is obtained according to the dynamic stress distribution map.

[0182] As a preferred solution, the floating platform life analysis system 300 further includes a target node analysis module, and the target node analysis module is used for:

[0183] Respond to the analysis instruction of the target node, and the analysis instruction includes a calculation accuracy requirement;

[0184] According to the analysis instruction, extract the data sequence corresponding to the position of the target node from the stress state data;

[0185] And based on the calculation accuracy requirement, analyze and obtain the stress analysis result of the target node based on the data sequence.

[0186] As a preferred solution, the life analysis module 305 identifies the fatigue cycle times of the key nodes according to the stress state data; and combines the preset material fatigue curve to obtain the remaining life of the key nodes, and then based on each of the remaining lives, obtains the fatigue life of the floating platform to be analyzed, including:

[0187] The life analysis module 305 divides the floating platform to be analyzed into multiple sub-units, and obtains the calculation accuracy of each sub-unit;

[0188] According to the calculation accuracy of each sub-unit, extract the stress data sequences of each key node from the stress state data;

[0189] Eliminate the noise of the stress data sequence through a low-pass filter to obtain denoised data; and perform a moving average process on the denoised data to obtain smoothed data;

[0190] According to the preset stress amplitude and stress mean value, perform statistics on the smoothed data to obtain the fatigue cycle times of each key node;

[0191] According to the fatigue cycle times and the material fatigue curve, combined with the preset single damage value, calculate the cumulative damage of the key nodes using the Miner linear cumulative damage theory;

[0192] Obtain the preset initial life of the key node, subtract the cumulative damage from the initial life, calculate the remaining life of the key node; based on each of the remaining lives, obtain the fatigue life of the floating platform to be analyzed.

[0193] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0194] The present invention application obtains the monitoring data of the wind field, wave field and current field in the ocean area where the floating platform to be analyzed is located, extracts the initial spatio-temporal sequence, and obtains the coupling characteristics, so as to determine the dynamic characteristics under the interaction of multiple factors of the wind field, wave field and current field; further, according to the resolution of the monitoring data at different positions in the ocean area where the floating platform to be analyzed is located, grids with different degrees of refinement are respectively generated. Relatively fine grids are generated at positions with relatively large resolution, and relatively sparse grids are generated at positions with relatively small resolution, obtaining a multi-resolution feature distribution, which can realize the targeted distribution of the resolution in different regions, realize the dynamic adaptation of the ocean area where the floating platform to be analyzed is located under the action of wind-wave-current coupling loads, and then obtain an accurate load spectrum, and on this basis, an effective balance between simulation accuracy and calculation cost is achieved; in addition, since the load spectrum is obtained through time series analysis of the multi-resolution feature distribution, the time evolution characteristics of the dynamic coupling loads of the wind field, wave field and current field can be extracted. When calculating the stress distribution of the key nodes of the floating platform to be analyzed, accurate stress distribution data can be obtained, and then the remaining life of the key nodes and the fatigue life of the floating platform can be predicted. Compared with the analysis of a single factor in the prior art, the present invention application considers the action of the dynamic coupling loads of wind-wave-current, reduces the deviation between the floating platform life analysis result and the actual service state, and improves the accuracy of the analysis result; finally, this application calculates the stress distribution of multiple key nodes to obtain the stress state data of the floating platform to be analyzed as a whole, and uses the stress state data of the floating platform with integrity and smoothness to identify the fatigue cycle times of the key nodes and then obtain the remaining life. In this way, compared with the method of directly predicting the remaining life of the key nodes using the stress distribution of the key nodes, it can avoid the excessive deviation of the remaining life identified by each key node, resulting in the distortion of the fatigue life data of the floating platform finally analyzed. That is, it is equivalent to performing a similar smoothing process using the stress state data of the floating platform to be analyzed as a whole before identifying the remaining life of each key node, controlling the remaining life of each key node within a relatively smaller range, so as to analyze and obtain a more accurate and stable fatigue life of the floating platform to be analyzed.

[0195] The specific embodiments described above further elaborate on the purpose, technical solutions and beneficial effects of the present invention. It should be understood that the above description is only the specific embodiments of the present invention and is not used to limit the protection scope of the present invention. In particular, it is pointed out that for those skilled in the art, any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A floating platform life analysis method based on wind, wave and current dynamic loads, characterized in that: include: Acquire the monitoring data of wind field, wave field and flow field of the ocean area where the floating platform to be analyzed is located in real time, and extract the initial time-space sequence from the monitoring data; wherein the floating platform to be analyzed includes a plurality of key nodes; Extract coupling features from the initial spatiotemporal sequence; based on the coupling features, generate grids with different degrees of refinement according to the resolution of the monitoring data at different locations in the ocean area where the floating platform to be analyzed is located, and obtain a multi-resolution feature distribution according to the grid generation results; wherein the grid corresponding to the area with relatively large resolution has a finer degree than the grid corresponding to the area with relatively small resolution; Analyzing the time series of the multi-resolution characteristic distribution to obtain a load spectrum; Based on the load spectrum, finite element analysis is used to calculate the stress distribution of the plurality of key nodes to obtain stress state data of the floating platform to be analyzed; According to the stress state data, the number of fatigue cycles of the key nodes is identified; and the remaining life of the key nodes is obtained in combination with the preset material fatigue curve, and then based on each of the remaining lives, the fatigue life of the floating platform to be analyzed is obtained; The degree of refinement is set according to the unit length, the number of grids and the grid density; Based on the coupling feature, according to the resolution of the monitoring data at different locations in the ocean area where the floating platform to be analyzed is located, grids with different degrees of refinement are generated respectively, and multi-resolution feature distribution is obtained according to the grid generation result, including: Dividing the ocean area where the floating platform to be analyzed is located into a plurality of sub-areas; Build multiple computing queues, each of which corresponds to a sub-region; Determine the unit length, number of grids and grid density of each sub-area according to the resolution of the monitoring data of each sub-area; Controlling the plurality of calculation queues to generate target grids for each sub-region according to the unit length, the number of grids and the grid density of each sub-region; Based on the coupling features and the target grids of each sub-region, the multi-resolution feature distribution is obtained; The controlling the plurality of calculation queues to generate target grids for each sub-region according to the unit length, the number of grids and the grid density of each sub-region respectively includes: Controlling the plurality of calculation queues to respectively preliminarily generate mesh surface information according to the unit length, mesh number and mesh density of each of the sub-regions; Controlling the plurality of calculation queues to generate three-dimensional grids of each sub-area based on the grid surface information and obtain grid body information; According to the mesh body information, the vertices and body units of each three-dimensional mesh are obtained; and the vertices and body units of each three-dimensional mesh are numbered to obtain a unique string of each vertex and a unique string of each body unit; Controlling the multiple computing queues to perform synchronous communication so that the body units and unique character strings between adjacent sub-regions are projected onto each other to obtain projection information; The three-dimensional grid of each sub-region is optimized according to the projection information to obtain the target grid of each sub-region.

2. A floating platform life analysis method based on wind, wave and current dynamic loads as claimed in claim 1, characterized in that: The step of extracting the initial spatiotemporal sequence from the monitoring data comprises: Performing standardization processing on the monitoring data of the wind field, wave field and flow field to obtain standardized data; A preset convolutional neural network is used to extract wind field features, wave field features and flow field features from the standardized data to obtain the initial space-time sequence.

3. A floating platform life analysis method based on wind, wave and current dynamic loads as claimed in claim 1, characterized in that: The step of extracting coupling features from the initial spatiotemporal sequence comprises: extracting a preliminary feature set of wind field, wave field and flow field coupling from the initial space-time sequence; When the feature dimension of the preliminary feature set is greater than a preset dimension threshold, performing dimensionality reduction processing on the preliminary feature set to obtain dimensionality reduction data; The time series of the dimension reduction data is analyzed, and the correlation features of the dynamic characteristics of the interaction between the wind field, the wave field and the flow field are enhanced to obtain the coupling features.

4. A floating platform life analysis method based on wind, wave and current dynamic loads as claimed in claim 1, characterized in that: The floating platform to be analyzed is connected to at least one anchor chain; the stress distribution of the plurality of key nodes is calculated by finite element analysis based on the load spectrum to obtain stress state data of the floating platform to be analyzed, including: Constructing a platform model of the floating platform to be analyzed; constructing anchor chain models of the anchor chains respectively, embedding all anchor chain models into the platform model, and obtaining a target model; Constructing external excitation based on the load spectrum; performing finite element analysis based on the target model and the external excitation in combination with preset boundary conditions to obtain stress distribution of each key node; According to the stress distribution of each key node, analyze the temporal and spatial variation trends of the stress of each key node; Based on the temporal variation trend and spatial variation trend of the stress of each of the key nodes, a dynamic stress distribution diagram of the floating platform to be analyzed is obtained, and the stress state data is obtained according to the dynamic stress distribution diagram.

5. The floating platform life analysis method based on wind, wave and current dynamic loads according to claim 1, characterized in that: The floating platform life analysis method further comprises: Responding to an analysis instruction of a target node, wherein the analysis instruction includes a calculation accuracy requirement; According to the analysis instruction, extracting a data sequence of a position corresponding to the target node from the stress state data; And according to the calculation accuracy requirement, the stress analysis result of the target node is obtained based on the data sequence analysis.

6. A floating platform life analysis system based on wind, wave and current dynamic loads, characterized in that: Used to implement a floating platform life analysis method based on wind, wave and current dynamic loads as described in any one of claims 1 to 5; the floating platform life analysis system includes a spatiotemporal sequence extraction module, a characteristic distribution generation module, a load spectrum analysis module, a stress acquisition module and a life analysis module; wherein, The spatiotemporal sequence extraction module is used to obtain in real time the monitoring data of the wind field, wave field and flow field of the ocean area where the floating platform to be analyzed is located, and extract the initial spatiotemporal sequence from the monitoring data; wherein the floating platform to be analyzed includes a plurality of key nodes; The feature distribution generation module is used to extract coupling features from the initial spatiotemporal sequence; based on the coupling features, grids with different degrees of refinement are generated according to the resolution of the monitoring data at different locations in the ocean area where the floating platform to be analyzed is located, and a multi-resolution feature distribution is obtained according to the grid generation results; wherein the grid corresponding to the area with relatively large resolution has a finer degree than the grid corresponding to the area with relatively small resolution; The load spectrum analysis module is used to analyze the time series of the multi-resolution feature distribution to obtain a load spectrum; The stress acquisition module is used to calculate the stress distribution of the plurality of key nodes by finite element analysis based on the load spectrum to obtain stress state data of the floating platform to be analyzed; The life analysis module is used to identify the number of fatigue cycles of key nodes according to stress state data; and to obtain the remaining life of key nodes in combination with preset material fatigue curves, and then to obtain the fatigue life of the floating platform to be analyzed based on each remaining life.

7. A floating platform life analysis system based on wind, wave and current dynamic loads as claimed in claim 6, characterized in that: The degree of refinement is set according to the unit length, the number of grids and the grid density; The feature distribution generation module generates grids of different degrees of refinement based on the coupling feature and according to the resolution of the monitoring data at different locations in the ocean area where the floating platform to be analyzed is located, and obtains a multi-resolution feature distribution according to the grid generation result, including: The characteristic distribution generating module divides the ocean area where the floating platform to be analyzed is located into a plurality of sub-areas; Build multiple computing queues, each of which corresponds to a sub-region; Determine the unit length, number of grids and grid density of each sub-area according to the resolution of the monitoring data of each sub-area; Controlling the plurality of calculation queues to generate target grids for each sub-region according to the unit length, the number of grids and the grid density of each sub-region; Based on the coupling features and the target grids of each sub-region, the multi-resolution feature distribution is obtained.

8. A floating platform life analysis system based on wind, wave and current dynamic loads as claimed in claim 7, characterized in that: The feature distribution generation module controls the multiple calculation queues to generate target grids for each sub-region according to the unit length, the number of grids and the grid density of each sub-region, including: The feature distribution generation module controls the multiple calculation queues to respectively preliminarily generate grid surface information according to the unit length, grid number and grid density of each sub-region; Controlling the plurality of calculation queues to generate three-dimensional grids of each sub-area based on the grid surface information and obtain grid body information; According to the mesh body information, the vertices and body units of each three-dimensional mesh are obtained; and the vertices and body units of each three-dimensional mesh are numbered to obtain a unique string of each vertex and a unique string of each body unit; Controlling the multiple computing queues to perform synchronous communication so that the body units and unique character strings between adjacent sub-regions are projected onto each other to obtain projection information; The three-dimensional grid of each sub-region is optimized according to the projection information to obtain the target grid of each sub-region.

Images