Floating platform fatigue damage assessment method and device

Abstract

The invention discloses a floating platform fatigue damage assessment method and a floating platform fatigue damage assessment device, which are used for solving the technical problem that the existing floating platform fatigue damage assessment method is difficult to comprehensively assess the fatigue damage of a floating platform by arranging a limited number of sensors. The method comprises the steps that a floating platform finite element model is established, a wave scatter diagram is obtained, and short-term working conditions corresponding to all local structures are determined through discrete long-term sea conditions; then, combining data such as preset rule waves and loading conditions, and identifying fatigue hotspots by adopting a hotspot analysis method; a hot spot stress prediction model is constructed by using a time domain finite element and machine learning; and finally, based on a rain flow counting method and a stress-life curve, generating a fatigue damage result of each fatigue hot spot.

Description

Technical Field

[0001] The present invention relates to the field of marine engineering technology, and in particular to a method and device for assessing fatigue damage of a floating platform. Background Art

[0002] Floating platforms are large structures used for marine development and operations, and are widely used in offshore oil and gas extraction, wind power generation, marine scientific research, and other fields. They offer numerous advantages, including strong adaptability, high cost-effectiveness, high safety, and mature technology. However, floating platforms often experience reciprocating motion due to the coupled effects of wind, waves, and currents, subjecting their structures to complex alternating loads. This can cause fatigue damage, impacting both structural safety and the safety of personnel.

[0003] Floating platforms usually operate in waters far from the coast, where the environmental loads are complex and they are continuously subjected to the effects of wind, waves and currents in the ocean. Their structures are subjected to alternating loads for a long time and are very prone to fatigue fracture. Therefore, it is very important to timely analyze and evaluate the fatigue damage of floating platform structures through on-site monitoring. This can obtain the safety status of the platform structure in real time and provide an important basis for efficient production, safe operation and timely maintenance.

[0004] Existing methods for fatigue damage assessment on floating platforms mostly deploy a limited number of sensors to monitor stress responses under different sea conditions and perform linear extrapolation to obtain hotspot stress data. Fatigue analysis is then performed using the rainflow counting method combined with the SN curve. However, floating platforms are large in size and have numerous fatigue hotspots. Furthermore, due to high deployment costs and limited physical space, some fatigue hotspots are not suitable for direct stress monitoring, making comprehensive fatigue damage assessment of floating platforms difficult. Summary of the Invention

[0005] The present invention provides a floating platform fatigue damage assessment method and device, which are used to solve the technical problem that most existing floating platform fatigue damage assessment methods use a limited number of sensors to achieve comprehensive assessment of the floating platform fatigue damage.

[0006] A first aspect of the present invention provides a method for fatigue damage assessment of a floating platform, comprising:

[0007] Obtaining a wave scatter diagram of the working sea area of the floating platform and establishing an overall finite element model of the floating platform corresponding to the floating platform;

[0008] Based on the wave scatter diagram and a plurality of preset wave spectra, discretizing the long-term sea conditions in the wave scatter diagram, and determining a plurality of short-term sea conditions corresponding to each local structure in the overall finite element model of the floating platform;

[0009] performing a fatigue hotspot analysis on the entire finite element model of the floating platform using a preset fatigue hotspot analysis method based on a plurality of preset regular wave sea conditions, the wave scatter diagram, a plurality of preset floating platform loading conditions, and a plurality of short-term sea conditions corresponding to each of the local structures, to determine a plurality of fatigue hotspots;

[0010] A fatigue hotspot stress response prediction model is constructed using a time-domain finite element method and a machine learning algorithm based on each fatigue hotspot, each preset floating platform loading condition, each preset wave spectrum, and multiple stress monitoring points of the floating platform;

[0011] Based on the three-point rainflow counting method and the time-domain finite element method, the fatigue hotspot stress response prediction model is used to generate fatigue damage results for each fatigue hotspot according to each stress monitoring point, the extreme environmental load conditions corresponding to the floating platform, multiple preset floating platform loading conditions, multiple short-term sea conditions corresponding to each local structure, and preset stress-life curves.

[0012] Optionally, the preset fatigue hotspot analysis method includes offshore mobile platform classification rules, stress response under hazardous conditions, and fatigue spectrum analysis methods; the preset fatigue hotspot analysis method is used to perform fatigue hotspot analysis on the entire finite element model of the floating platform according to multiple preset regular wave sea conditions, the wave scatter diagram, multiple preset floating platform loading conditions, and multiple short-term sea conditions corresponding to each of the local structures, to determine multiple fatigue hotspots, including:

[0013] Based on the offshore mobile platform classification rules, analyzing the overall finite element model of the floating platform to identify multiple fatigue hotspots;

[0014] Analyzing the overall finite element model of the floating platform based on the stress response under the hazardous working condition to identify multiple fatigue hotspots;

[0015] The fatigue spectrum analysis method is used to analyze the overall finite element model of the floating platform according to the multiple preset regular wave sea conditions, the wave scatter diagram, multiple preset floating platform loading conditions and multiple short-term sea conditions corresponding to each of the local structures to determine multiple fatigue hotspots.

[0016] Optionally, the fatigue spectrum analysis method is used to analyze the entire finite element model of the floating platform according to the plurality of preset regular wave sea conditions, the wave scatter diagram, a plurality of preset floating platform loading conditions, and a plurality of short-term sea conditions corresponding to each of the local structures to determine a plurality of fatigue hotspots, including:

[0017] constructing stress transfer functions for each local structure of the overall finite element model of the floating platform under each of the preset regular wave sea conditions and each of the preset floating platform loading conditions, and generating a plurality of stress transfer functions corresponding to each of the local structures;

[0018] determining, based on a plurality of short-term sea conditions corresponding to each of the local structures, a wave spectrum of each of the local structures under each of the short-term sea conditions;

[0019] Calculating the stress energy spectrum density of each local structure under each of the short-term sea conditions based on a plurality of stress transfer functions corresponding to each of the local structures and a wave spectrum of each of the local structures under each of the short-term sea conditions;

[0020] Determining the joint probability of significant wave height and zero-crossing period for each of the short-term sea conditions based on the wave scatter diagram;

[0021] Calculating the total fatigue life of each of the local structures according to the stress energy spectrum density of each of the local structures under each of the short-term sea conditions, the significant wave height of each of the short-term sea conditions, and the joint probability of the zero-crossing period;

[0022] The total fatigue life of each of the local structures is sorted in ascending order, and the local structure corresponding to the total fatigue life of the first preset number of digits is selected as the fatigue hotspot.

[0023] Optionally, the use of a time-domain finite element method and a machine learning algorithm to construct a fatigue hotspot stress response prediction model based on each fatigue hotspot, each preset floating platform loading condition, each preset wave spectrum, and multiple stress monitoring points of the floating platform includes:

[0024] outputting simulated stress response time history data of the plurality of stress monitoring points and the simulated stress response time history data of the plurality of fatigue hotspots based on the preset floating platform loading conditions, the wave characteristic parameters corresponding to the preset wave spectra, the positions of the plurality of stress monitoring points and the plurality of fatigue hotspots using a time-domain finite element method;

[0025] Based on a machine learning algorithm, a fatigue hotspot stress response prediction model is constructed, which takes the simulated stress response time history data of multiple stress monitoring points as input and takes the simulated stress response time history data of multiple fatigue hotspots as output.

[0026] Optionally, the fatigue damage results include fatigue damage occurring each year and remaining fatigue life; the fatigue damage results of each fatigue hotspot are generated by using the fatigue hotspot stress response prediction model based on the three-point rainflow counting method and the time-domain finite element method according to each stress monitoring point, the extreme environmental load condition corresponding to the floating platform, multiple preset floating platform loading conditions, multiple short-term sea conditions corresponding to each local structure, and preset stress-life curves, including:

[0027] Based on the three-point rainflow counting method, the fatigue hotspot stress response prediction model is used to calculate the fatigue damage of each fatigue hotspot within the monitoring time according to the monitored stress response time history data corresponding to each stress monitoring point and the preset stress-life curve;

[0028] Calculating fatigue damage caused by long-term environmental loads at each fatigue hotspot using the three-point rainflow counting method and the time-domain finite element method based on the preset stress-life curve, multiple preset floating platform loading conditions, and multiple short-term sea conditions corresponding to each local structure;

[0029] The three-point rainflow counting method and the time-domain finite element method are used to calculate the fatigue damage of each fatigue hotspot under extreme working conditions according to the extreme environmental load conditions corresponding to the floating platform, a plurality of preset floating platform loading conditions, and the preset stress-life curve;

[0030] The fatigue damage of each fatigue hotspot during the monitoring period, the fatigue damage caused by long-term environmental loads, and the fatigue damage under extreme working conditions are respectively added to determine the cumulative fatigue damage of each fatigue hotspot;

[0031] Based on the accumulated fatigue damage of each fatigue hotspot, the fatigue damage occurring each year and the remaining fatigue life of each fatigue hotspot are calculated.

[0032] Optionally, the three-point rainflow counting method is based on the fatigue hotspot stress response prediction model, and the fatigue damage within the monitoring time of each fatigue hotspot is calculated according to the monitored stress response time history data corresponding to each stress monitoring point and the preset stress-life curve, including:

[0033] using the monitored stress response time history data corresponding to each of the stress monitoring points as input to the fatigue hotspot stress response prediction model, and outputting first stress response time history data of a plurality of fatigue hotspots;

[0034] Analyzing the first stress time history data of the plurality of fatigue hot spots using the three-point rainflow counting method to determine a plurality of first stress cycle amplitudes and the number of stress range cycles corresponding to each of the first stress cycle amplitudes;

[0035] determining, based on the preset stress-life curve and a plurality of the first stress cycle amplitudes, an average number of times fatigue failure of the structure occurs corresponding to each of the first stress cycle amplitudes;

[0036] The fatigue damage of each fatigue hotspot within the monitoring time is calculated according to the average number of times fatigue failure of the structure occurs and the number of stress range cycles corresponding to each of the first stress cycle amplitudes.

[0037] Optionally, the using the three-point rainflow counting method and the time-domain finite element method to calculate the fatigue damage caused by the long-term environmental load of each fatigue hotspot according to the preset stress-life curve, a plurality of preset floating platform loading conditions, and a plurality of short-term sea conditions corresponding to each local structure includes:

[0038] Determining the second stress time history data of the plurality of fatigue hot spots using the time domain finite element method according to the plurality of short-term sea conditions corresponding to the respective local structures and the plurality of preset floating platform loading conditions;

[0039] Analyzing the second stress time history data of the plurality of fatigue hot spots using the three-point rainflow counting method to determine a plurality of second stress cycle amplitudes and the number of stress range cycles corresponding to each of the second stress cycle amplitudes;

[0040] determining, based on the preset stress-life curve and a plurality of second stress cycle amplitudes, the number of stress cycles required for fatigue failure of the structure corresponding to each second stress cycle amplitude;

[0041] The fatigue damage caused by the long-term environmental load of each fatigue hotspot is calculated based on the number of stress cycles and the number of stress range cycles required for the structure to undergo fatigue failure corresponding to each second stress cycle amplitude.

[0042] Optionally, the using the three-point rainflow counting method and the time-domain finite element method to calculate the fatigue damage of each fatigue hotspot under extreme conditions according to the extreme environmental load conditions corresponding to the floating platform, a plurality of preset floating platform loading conditions, and the preset stress-life curve includes:

[0043] Determining the third stress time history data of the plurality of fatigue hot spots using the time domain finite element method according to the extreme environmental load conditions corresponding to the floating platform and the plurality of preset floating platform loading conditions;

[0044] Analyzing the third stress time history data of the plurality of fatigue hot spots using the three-point rainflow counting method to determine a plurality of third stress cycle amplitudes and the number of stress range cycles corresponding to each third stress cycle amplitude;

[0045] determining, based on the preset stress-life curve and the plurality of third stress cycle amplitudes, the number of stress cycles required for fatigue failure of the structure corresponding to each of the third stress cycle amplitudes;

[0046] The fatigue damage of each fatigue hotspot under extreme working conditions is calculated based on the number of stress cycles and the number of stress range cycles required for fatigue failure of the structure corresponding to each of the third stress cycle amplitudes.

[0047] A second aspect of the present invention provides a floating platform fatigue damage assessment device, comprising:

[0048] An acquisition module is used to obtain a wave scatter diagram of the working sea area of the floating platform and establish an overall finite element model of the floating platform corresponding to the floating platform;

[0049] a determination module, configured to discretize the long-term sea conditions in the wave scatter diagram based on the wave scatter diagram and a plurality of preset wave spectra, and determine a plurality of short-term sea conditions corresponding to each local structure in the overall finite element model of the floating platform;

[0050] an analysis module, configured to perform fatigue hotspot analysis on the entire finite element model of the floating platform using a preset fatigue hotspot analysis method according to a plurality of preset regular wave sea conditions, the wave scatter diagram, a plurality of preset floating platform loading conditions, and a plurality of short-term sea conditions corresponding to each of the local structures, to determine a plurality of fatigue hotspots;

[0051] a construction module for constructing a fatigue hotspot stress response prediction model based on each fatigue hotspot, each preset floating platform loading condition, each preset wave spectrum, and a plurality of stress monitoring points of the floating platform using a time-domain finite element method and a machine learning algorithm;

[0052] An evaluation module is configured to generate fatigue damage results for each fatigue hotspot based on a three-point rainflow counting method and the time-domain finite element method, using the fatigue hotspot stress response prediction model according to each stress monitoring point, the extreme environmental load conditions corresponding to the floating platform, multiple preset floating platform loading conditions, multiple short-term sea conditions corresponding to each local structure, and preset stress-life curves.

[0053] A third aspect of the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of any one of the above-described floating platform fatigue damage assessment methods.

[0054] It can be seen from the above technical solutions that the present invention has the following advantages:

[0055] The above technical solution of the present invention provides a method for fatigue damage assessment of a floating platform. First, a wave scatter diagram of the working sea area of the floating platform is obtained, and a floating platform overall finite element model corresponding to the floating platform is established; then, based on the wave scatter diagram and multiple preset wave spectra, the long-term sea conditions in the wave scatter diagram are discretized to determine multiple short-term sea conditions corresponding to each local structure in the overall finite element model of the floating platform; a preset fatigue hotspot analysis method is used to perform fatigue hotspot analysis on the overall finite element model of the floating platform according to multiple preset regular wave sea conditions, wave scatter diagrams, multiple preset floating platform loading conditions and multiple short-term sea conditions corresponding to each local structure to determine multiple fatigue hotspots; a time domain finite element method and a machine learning algorithm are used to analyze the fatigue hotspots, each preset wave condition, and multiple short-term sea conditions corresponding to each local structure. A fatigue hotspot stress response prediction model is constructed based on the floating platform loading conditions, various preset wave spectra, and multiple stress monitoring points of the floating platform. Finally, based on the three-point rainflow counting method and the time-domain finite element method, the fatigue hotspot stress response prediction model is used to generate fatigue damage results for each fatigue hotspot according to each stress monitoring point, the extreme environmental load conditions corresponding to the floating platform, multiple preset floating platform loading conditions, multiple short-term sea conditions corresponding to each local structure, and preset stress-life curves. Based on the above scheme, a fatigue hotspot stress response prediction model is established using a machine learning algorithm. Combined with the three-point rainflow counting method and the preset stress-life curve, the fatigue damage results for each fatigue hotspot in the floating platform are generated, thereby achieving a comprehensive assessment of the fatigue damage of the floating platform. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0057] Figure 1 A flowchart of a method for fatigue damage assessment of a floating platform provided in Example 1 of the present invention;

[0058] Figure 2 A schematic diagram of grid division of a floating platform provided in the first embodiment of the present invention;

[0059] Figure 3 A schematic diagram of the distribution of fatigue hot spots and stress monitoring points provided in Example 1 of the present invention;

[0060] Figure 4 A schematic diagram of the distribution of fatigue hot spots and stress monitoring points provided in Example 1 of the present invention;

[0061] Figure 5 A schematic diagram of a machine learning framework provided in Example 1 of the present invention;

[0062] Figure 6 A schematic diagram of fatigue hot spot stress response prediction data provided in Example 1 of the present invention;

[0063] Figure 7 A schematic diagram of fatigue hot spot stress prediction error analysis provided in Example 1 of the present invention;

[0064] Figure 8 A schematic diagram of a training loss curve provided in Example 1 of the present invention;

[0065] Figure 9 A schematic diagram of fatigue hot spot stress prediction value Vs actual value provided in the first embodiment of the present invention;

[0066] Figure 10 A schematic flow chart of a method for fatigue damage assessment of a floating platform provided in the first embodiment of the present invention;

[0067] Figure 11 This is a structural block diagram of a floating platform fatigue damage assessment device provided in Example 2 of the present invention. DETAILED DESCRIPTION

[0068] The embodiments of the present invention provide a floating platform fatigue damage assessment method and device, which are used to solve the technical problem that most existing floating platform fatigue damage assessment methods rely on the deployment of a limited number of sensors, making it difficult to achieve a comprehensive assessment of the fatigue damage of the floating platform.

[0069] In order to make the purpose, features, and advantages of the present invention more obvious and easy to understand, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described below are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0070] See also Figure 1 , Figure 1 This is a flowchart of the steps of a floating platform fatigue damage assessment method provided in Example 1 of the present invention.

[0071] The present invention provides a floating platform fatigue damage assessment method, comprising:

[0072] Step 101: Obtain a wave scatter diagram of the working sea area of the floating platform, and establish an overall finite element model of the floating platform corresponding to the floating platform.

[0073] It should be noted that the established finite element model of the floating platform as a whole includes a wetted surface model, a structural model, a mass model, and a Morrison model. The wetted surface model describes the surface of the floating platform in contact with the water and is primarily used to calculate the hydrodynamic loads during wave-structure interaction. The structural model analyzes the displacement, stress, and strain of the floating platform and is key to assessing the platform's structural integrity and safety. The mass model simulates the mass distribution of the floating platform, and the Morrison model calculates wave forces on slender structures (such as buoys, columns, and cross braces).

[0074] For example, assuming that the floating platform is 91.6m long, 76.0m wide, and weighs 26039t, Figure 2 As shown in the figure, the wet surface model, cross-bracing Morrison model, structural model and mass model of the floating platform are established, and the finite element mesh size is divided into 1m, a total of 32404 units and 28919 nodes. The mesh division is as follows: Figure 2 The main dimensions of the floating platform are shown in Table 1:

[0075] Table 1 Main dimensions of floating platform

[0076]

[0077] Step 102: Based on the wave scatter diagram and a plurality of preset wave spectra, the long-term sea conditions in the wave scatter diagram are discretized to determine a plurality of short-term sea conditions corresponding to the local structures in the overall finite element model of the floating platform.

[0078] The wave scatter diagram is a wave scatter diagram of the working sea area where the floating platform is located.

[0079] It should be noted that, based on the wave scatter diagram of the working sea area of the floating platform and the selected wave spectrum (i.e., preset wave spectrum), such as the JONSWAP spectrum, the Pierson-Moskowitz spectrum, the Bretschneider spectrum, etc., the long-term sea conditions in the wave scatter diagram are discretized into a series of short-term sea conditions (i.e., multiple short-term sea conditions corresponding to each local structure in the overall finite element model of the floating platform).

[0080] Step 103: Using a preset fatigue hotspot analysis method, fatigue hotspot analysis is performed on the entire finite element model of the floating platform according to multiple preset regular wave sea conditions, wave scatter diagrams, multiple preset floating platform loading conditions, and multiple short-term sea conditions corresponding to each local structure to determine multiple fatigue hotspots.

[0081] The preset fatigue hotspot analysis method includes the classification rules for offshore mobile platforms, stress response under hazardous conditions, and fatigue spectrum analysis method.

[0082] It should be noted that fatigue hotspots are areas within a structure where fatigue cracks are likely to initiate and propagate due to factors such as geometric discontinuities, stress concentrations, or material properties. These areas typically experience high stress levels and are at high risk for structural fatigue failure. For the floating platform being assessed, fatigue hotspots are identified based on the classification rules for mobile offshore platforms, stress responses under hazardous conditions, and fatigue spectrum analysis methods.

[0083] Specifically, step 103 may include the following sub-steps S31-S33:

[0084] Step S31: Analyze the overall finite element model of the floating platform based on the classification rules for offshore mobile platforms to determine multiple fatigue hotspots;

[0085] The Offshore Mobile Platform Classification Rules are a set of technical standards and rules for offshore mobile platforms (such as floating platforms). Their core purpose is to ensure the safety, reliability, and structural integrity of offshore mobile platforms.

[0086] It should be noted that according to the classification rules for mobile offshore platforms, fatigue design should be performed for any welds that could potentially develop fatigue cracks and any structural configuration that could cause stress concentration. Fatigue analysis of joint details should be performed, if necessary. For surface platforms (i.e., floating platforms), this includes at least the following: the connections between longitudinals and transverse webs and transverse bulkheads; the connections between transverse webs and shell plating, bottom plating, inner bottom plating, and bulkhead plating; the connections between the hopper tank sloping plating and the inner bottom plating and longitudinal bulkhead plating; and the connections between the end brackets of longitudinal and transverse girders. Therefore, these locations are identified as fatigue hotspots.

[0087] Step S32: Analyze the overall finite element model of the floating platform based on the stress response under the dangerous working condition to determine multiple fatigue hot spots;

[0088] Stress response under hazardous operating conditions: For structures such as floating platforms, hazardous operating conditions refer to those operating conditions that pose a threat to structural safety, such as maximum lateral force, maximum torsional force, maximum longitudinal shear force, and maximum vertical bending. The stress response refers to the stresses experienced by the structure under these hazardous conditions. Fatigue hotspots are typically areas of concentrated fatigue stress, and the stress response under hazardous conditions can indicate a structure's susceptibility to fatigue failure.

[0089] It should be noted that fatigue hotspots are typically areas of fatigue stress concentration, so the stress response under hazardous conditions can reflect a structure's susceptibility to fatigue failure. Stress concentration refers to the phenomenon in a structure where local stresses are significantly higher than the average stress due to geometric discontinuities or non-uniform material properties. This can be measured using the stress concentration factor. ,in is the stress concentration factor, is the local maximum stress, is the mean stress.

[0090] Furthermore, according to the recommendations of DNV (Det Norske Veritas), the maximum stress and stress concentration area of the platform under the maximum lateral stress state, maximum torsional stress state, maximum longitudinal shear state and maximum vertical bending state of the floating platform were checked, and this location was identified as a fatigue hotspot.

[0091] Furthermore, the maximum lateral stress state occurs when the floating platform is subjected to waves with a 90-degree wave direction and a wavelength twice the distance between the outer edges of the two buoys. In this state, the cross braces are subject to significant lateral separation and lateral compression forces. The maximum torsional stress state occurs when the floating platform is subjected to waves with a 45-degree wave direction and a wavelength of one diagonal length. In this state, not only torque but also separation forces act on the structure. The maximum longitudinal shear state occurs when the floating platform is subjected to waves with a 45-degree wave direction and a wavelength of 1.5 diagonal lengths. When the phase is 180 degrees, the longitudinal shear force may cause high shear stresses in the platform's connection nodes and at the connections between columns and buoys. These areas are often fatigue hotspots. The maximum vertical bending state occurs when the floating platform is subjected to waves with a 0-degree wave direction and a wavelength equal to the length of the buoys. When the wave phase is 0 degrees, the center of the platform is at the wave crest, and the bow and stern are at the wave trough. At this time, the buoyancy in the center is greater than the weight, while the buoyancy at the bow and stern is less than the weight, causing the structure to arch. When the wave phase is 180 degrees, similarly, the structure will sag.

[0092] Therefore, the above four regular wave conditions are loaded onto the floating platform to obtain the stress response of the floating platform under dangerous conditions, and the maximum stress and stress concentration points under each condition are identified as fatigue hotspots.

[0093] For example, see Figure 3-Figure 4 , combined with the offshore platform classification rules, stress response under dangerous conditions and fatigue spectrum analysis results to determine the fatigue hot spots. The distribution of fatigue hot spots and stress monitoring points is as follows Figure 3-Figure 4 The fatigue damage degree of fatigue hot spots and stress monitoring points is shown in Table 2:

[0094] Table 2 Fatigue hot spots and fatigue damage degree of stress monitoring points

[0095]

[0096] Step S33: using a fatigue spectrum analysis method to analyze the overall finite element model of the floating platform according to multiple preset regular wave sea conditions, wave scatter diagrams, multiple preset floating platform loading conditions, and multiple short-term sea conditions corresponding to each local structure, to determine multiple fatigue hotspots.

[0097] The fatigue spectrum analysis method is a frequency-domain-based fatigue life assessment method with low computational complexity and high accuracy, which can basically meet engineering needs. It is especially suitable for structural fatigue analysis under complex working conditions and random loads.

[0098] Specifically, step S33 may include the following sub-steps S331-S336:

[0099] Step S331: constructing stress transfer functions for each local structure of the overall finite element model of the floating platform under each preset regular wave sea condition and each preset floating platform loading condition, thereby generating a plurality of stress transfer functions corresponding to each local structure;

[0100] It should be noted that a series of regular wave sea conditions are preset (i.e., multiple preset regular wave sea conditions), multiple preset floating platform loading conditions, determine the stress transfer function of each local structure of the overall finite element model of the floating platform under a series of regular wave sea conditions under different loading conditions (preset floating platform loading conditions), that is, multiple stress transfer functions corresponding to each local structure .in is the wave direction angle, is the wave frequency.

[0101] Step S332: determining the wave spectrum of each local structure under each short-term sea condition based on the multiple short-term sea conditions corresponding to each local structure;

[0102] It should be noted that, based on the wave scatter diagram of the working area of the floating platform and the selected wave spectra, such as the JONSWAP spectrum, the Pierson-Moskowitz spectrum, the Bretschneider spectrum, etc., the long-term sea state is discretized into a series of short-term sea conditions, and then the wave spectrum of each local structure under each short-term sea condition is obtained. ,in For the meaningful wave height, It is a zero-crossing period.

[0103] Step S333: Calculate the stress energy spectrum density of each local structure under each short-term sea condition based on the multiple stress transfer functions corresponding to each local structure and the wave spectrum of each local structure under each short-term sea condition;

[0104] It should be noted that according to the stress transfer function , wave spectrum , the stress energy spectrum density of each local structure under each short-term sea condition , the calculation formula of stress energy spectrum density is:

[0105] ;

[0106] Step S334: determining the joint probability of the significant wave height and the zero-crossing period for each short-term sea condition based on the wave scatter diagram;

[0107] It should be noted that the wave scatter diagram is discretized into a series of short-term sea conditions according to the selected wave spectrum. The results of the short-term sea conditions are the significant wave height, the zero-crossing period and the probability of the two, that is, the joint probability p of the significant wave height and the zero-crossing period of each short-term sea condition i .

[0108] Step S335: Calculate the total fatigue life of each local structure based on the stress energy spectrum density of each local structure under each short-term sea state, the significant wave height of each short-term sea state, and the joint probability of the zero-crossing period;

[0109] It should be noted that it is assumed that the probability density function of the short-term distribution of stress range conforms to the Rayleigh distribution. Applying the obtained spectral moment, the average upper zero-crossing period f and the spectral width parameter of the probability density function of the short-term distribution of stress range are It can be calculated as follows:

[0110] Rayleigh distribution probability density function :

[0111] ;

[0112] Average zero crossing period:

[0113] ;

[0114] Spectral width parameters:

[0115] ;

[0116] Where S is the stress range (twice the stress amplitude), , 、 、 are the zero-order moment, second-order moment, and fourth-order moment of the stress spectrum, respectively.

[0117] Specifically, the stress energy spectrum density is first used to calculate the spectral moment of the stress energy spectrum. The calculation process is as follows:

[0118] ;

[0119] in, is the spectral moment of the n-th order stress energy spectrum; Frequency The nth power, n takes the corresponding value (2 / 0).

[0120] Next, the upper zero-crossing frequency of the stress response is calculated using the spectral moment of the stress energy spectrum. , and calculate the Rayleigh distribution probability density ,in, is the zero-crossing frequency of the stress response of the local structure in the i-th short-term sea state; is the spectral moment of the second-order stress energy spectrum of the local structure in the i-th short-term sea state, that is, the second-order moment of the stress spectrum of the local structure in the i-th short-term sea state; is the spectral moment of the zero-order stress energy spectrum of the local structure in the i-th short-term sea state, that is, the zero-order moment of the stress spectrum of the local structure in the i-th short-term sea state; is the Rayleigh distribution probability density of the local structure under the i-th short-term sea state; is the standard deviation of the stress of the local structure under the i-th short-term sea state, which is used to characterize the degree of dispersion of the stress range or the width of the distribution. It reflects the fluctuation of the stress amplitude. Indicates that the difference in stress amplitude is large, and the smaller It means that the stress amplitude is relatively concentrated. ; is the stress range (twice the stress amplitude).

[0121] Furthermore, the short-term fatigue damage caused by each local structure under each short-term sea condition is calculated by using the zero-crossing frequency of the stress response, the probability density of the Rayleigh distribution, the significant wave height and the joint probability of the zero-crossing period. Then, the short-term fatigue damage caused by each local structure under each short-term sea condition is accumulated respectively to obtain the cumulative damage of each local structure. Specifically, the present invention applies the Palmgren-Miner linear cumulative damage theory to calculate the cumulative fatigue damage. When the probability density function of the short-term distribution of the stress range generated by a certain short-term sea condition can be expressed by the Rayleigh distribution, it is assumed that the form of the SN curve is N=KS -m , then the short-term fatigue damage caused by the i-th short-term sea condition, that is, the short-term fatigue damage caused by the local structure under the i-th short-term sea condition is:

[0122] ;

[0123] in, is the short-term fatigue damage caused to the local structure under the i-th short-term sea condition; T is the design life; K and m are two physical parameters that define the SN curve; is the upper zero-crossing frequency of the stress response of the local structure under the i-th short-term sea state, that is, the average action frequency of the stress range, in Hertz; is the joint probability of the significant wave height and the zero-crossing period of the i-th short-term sea state; is the Rayleigh distribution probability density of the local structure in the i-th short-term sea state, which represents the probability density of the local structure generating the S value in the i-th short-term sea state; S is the stress range, In a short-term sea state, many stress amplitudes will appear. Since the values are different, there are m amplitudes, and the probability of each is 1 / m. However, many stress amplitudes are similar, so similar stress amplitudes are merged, and the number of stress amplitudes is much smaller. For example, the stress amplitude of [10, 11] MPa is considered to be 10.5 MPa, and the m stress amplitudes are calculated separately by default.

[0124] Furthermore, the damage caused by the M short-term sea conditions in the wave scatter diagram is accumulated to obtain the total cumulative damage D, that is, the cumulative damage of each local structure. The calculation process can be expressed as:

[0125] ;

[0126] Where D is the cumulative damage of the local structure; is the “average” frequency of the stress range S at the calculation point over the entire life of the structure, in Hertz, ; M is the total number of short-term sea conditions.

[0127] Finally, the total fatigue life of each local structure is calculated using the cumulative damage of each local structure, where the total fatigue life is The calculation process is as follows:

[0128] ;

[0129] It is worth mentioning that the present invention introduces the probability density function of stress range distribution and the total number of cycles during the structural design life N T , we can get:

[0130] ;

[0131] Based on the above foundation, the total cumulative fatigue damage D can also be written as:

[0132] ;

[0133] in, is the mth power of the stress range S; m is the physical parameter in the SN curve.

[0134] Step S336 : sorting the total fatigue life of each local structure in ascending order, and selecting the local structure corresponding to the total fatigue life of the first preset number of digits as the fatigue hotspot.

[0135] It should be noted that based on the above steps, the fatigue spectrum analysis method is used to obtain the total fatigue life of each local part of the floating platform for the established overall finite element model of the floating platform. The total fatigue life is sorted from short to long (small to large), and a series of the shortest points are selected (such as the first 5%, which can be adjusted according to the complexity of the structure and the amount of calculation). That is, the local structure corresponding to the total fatigue life of the first preset number of digits is determined as the fatigue hotspot.

[0136] In summary, fatigue hotspots identified based on offshore platform classification rules, hazardous operating condition stress responses, and fatigue spectrum analysis methods were statistically analyzed and the union of these three was taken to ultimately determine a total of N fatigue hotspots. Furthermore, M stress monitoring points were determined based on the current installation status of the floating platform's stress monitoring devices.

[0137] Step 104: Using a time-domain finite element method and a machine learning algorithm, a fatigue hotspot stress response prediction model is constructed based on each fatigue hotspot, each preset floating platform loading condition, each preset wave spectrum, and multiple stress monitoring points of the floating platform.

[0138] Specifically, step 104 may include the following sub-steps S41-S42:

[0139] Step S41: using a time-domain finite element method to output simulated stress response time history data of multiple stress monitoring points and simulated stress response time history data of multiple fatigue hotspots based on each preset floating platform loading condition, wave characteristic parameters corresponding to each preset wave spectrum, and the positions of multiple stress monitoring points and multiple fatigue hotspots;

[0140] Step S42: constructing a fatigue hotspot stress response prediction model based on a machine learning algorithm, which takes the simulated stress response time history data of multiple stress monitoring points as input and takes the simulated stress response time history data of multiple fatigue hotspots as output.

[0141] It should be noted that, with the wave characteristic parameters corresponding to the floating platform loading condition and the preset wave spectrum as input parameters, the time-domain finite element method is used to calculate the structural stress response under various combinations corresponding to the input parameters, and the simulated stress response time-history data of multiple stress monitoring points and the simulated stress response time-history data of multiple fatigue hotspots are output, so as to establish a stress response time-history database of stress monitoring points and fatigue hotspots, and a machine learning algorithm is used to construct a prediction model (fatigue hotspot stress response prediction model) with the monitoring point stress response (simulated stress response time-history data of stress monitoring points) as input and the fatigue hotspot stress response (simulated stress response time-history data of fatigue hotspots) as output.

[0142] Furthermore, in this step, a series of conditions including platform loading conditions, wave direction, etc. are determined according to the characteristics of the operating sea area. 、Yiyi Bogao , spectral peak period , wave spectrum. The floating platform loading conditions include typical conditions such as self-propelled conditions, operating conditions, and survival conditions; the wave spectrum includes typical wave spectrum types such as JONSWAP spectrum, Pierson-Moskowitz spectrum, and Bretschneider spectrum. The corresponding wave spectrum is selected according to the operating sea area and the applicable range of the wave spectrum; significant wave height Starting from 1m, with a step length of 0.5m, to the design survival limit state wave height; spectrum peak period Starting from 2s, with a step length of 0.5s, and ending at 30s; Starting from 0°, with a step size of 5°, and ending at 175°, the finite element simulation only needs to be performed for waves with a wave direction of [0,180°] and a step size of 5° due to the symmetry of the floating platform.

[0143] Furthermore, this step constructs a fatigue hotspot stress prediction model system based on machine learning. Targeting the spatial coupling characteristics of offshore platform structures, an adaptive neural network architecture driven by correlation analysis is employed. Feature correlation analysis and data distribution assessment are used to determine the spatial configuration and scale of input layer monitoring nodes. The model architecture is based on a multi-layer perceptron (MLP) and establishes a nonlinear mapping relationship between monitoring data and the target stress field through hierarchical feature transformation. The input layer quantitatively selects time-series stress data from key monitoring points based on spatial correlation metrics. During model construction, an adaptive gradient optimization algorithm dynamically adjusts learning parameters, combining early stopping with regularization to balance model complexity and generalization capability. Furthermore, for each fatigue hotspot, several nearest monitoring points are selected as input points, and a dependency analysis is performed on the number of monitoring points to determine the optimal number of monitoring points for the target prediction accuracy. During the performance evaluation phase, a multi-dimensional validation system is established, using three metrics—the coefficient of determination (R²), mean absolute error (MAE), and mean squared error (MSE)—to quantitatively evaluate prediction accuracy, error stability, and generalization performance. This method achieves effective prediction of the stress field of complex marine structures through the collaborative optimization of data-driven and mechanism-constrained methods.

[0144] For example, a fatigue hotspot stress response prediction model based on a machine learning algorithm generates a fatigue hotspot stress response database from 6 stress monitoring points and fatigue hotspot stress time history data. The stress response data of the 6 stress monitoring points are used as the input layer, and the model includes two fully connected hidden layers. The first hidden layer is configured with 20 neurons, and the second hidden layer is configured with 10 neurons. The machine learning framework is as follows: Figure 5 The stress response prediction data of fatigue hotspot is shown in Figure 6 As shown, according to Figure 6-Figure 7The prediction error of the model on the validation set has steadily converged to within 8%, the mean absolute error (MAE) has reached 0.89MPa, and the correlation coefficient (R²) has exceeded 0.91. Figure 8 The training loss curve shown verifies the effective convergence characteristics of the model. Figure 9 The prediction-measurement comparison diagram intuitively demonstrates the high-precision prediction capability of the present invention in complex marine environments.

[0145] Step 105: Based on the three-point rainflow counting method and the time-domain finite element method, a fatigue hotspot stress response prediction model is used to generate fatigue damage results for each fatigue hotspot according to each stress monitoring point, the extreme environmental load conditions corresponding to the floating platform, multiple preset floating platform loading conditions, multiple short-term sea conditions corresponding to each local structure, and preset stress-life curves.

[0146] Fatigue damage results include fatigue damage occurrence per year and remaining fatigue life.

[0147] It should be noted that the cumulative fatigue damage of the fatigue hotspot is composed of the fatigue damage during the monitoring period, the fatigue damage caused by long-term environmental loads during the service period, and the damage caused by extreme environmental loads. The three parts of fatigue damage and the annual fatigue damage of the fatigue hotspot are calculated separately to finally obtain the remaining fatigue life of the fatigue hotspot.

[0148] Specifically, step 105 may include the following sub-steps S51-S55:

[0149] Step S51: Based on the three-point rainflow counting method, a fatigue hotspot stress response prediction model is used to calculate the fatigue damage of each fatigue hotspot within the monitoring time according to the monitored stress response time history data corresponding to each stress monitoring point and a preset stress-life curve;

[0150] Furthermore, step S51 may include the following sub-steps S511-S514:

[0151] Step S511: using the monitored stress response time history data corresponding to each stress monitoring point as input to a fatigue hotspot stress response prediction model, and outputting first stress response time history data of multiple fatigue hotspots;

[0152] Step S512: Analyze the first stress time history data of multiple fatigue hot spots using a three-point rainflow counting method to determine multiple first stress cycle amplitudes and the number of stress range cycles corresponding to each first stress cycle amplitude;

[0153] Step S513: determining the average number of fatigue failures of the structure corresponding to each first stress cycle amplitude based on the preset stress-life curve and the plurality of first stress cycle amplitudes;

[0154] Step S514: Calculate the fatigue damage of each fatigue hotspot within the monitoring time according to the average number of times the structure fatigue failure occurs and the number of stress range cycles corresponding to each first stress cycle amplitude.

[0155] It should be noted that the stress time history data for the corresponding monitoring points (i.e., the monitored stress response time history data corresponding to each stress monitoring point) is input, and the stress time history of N fatigue hotspots (i.e., the first stress response time history data of multiple fatigue hotspots) is obtained using the constructed fatigue hotspot stress response prediction model. The fatigue damage of each fatigue hotspot during the stress monitoring period is calculated using the rain flow counting method and a selected SN curve (preset stress-life curve). Preferably, fatigue damage calculation is performed using the 3-hour stress prediction data from the stress monitoring point, with the fatigue damage value continuously updated.

[0156] Specifically, the three-point rain flow counting method is used to analyze the fatigue hotspot stress prediction data (the first stress time history data of the fatigue hotspot). For the k-th fatigue hotspot, m groups of data with different stress cycle amplitudes are obtained (i.e., multiple first stress cycle amplitudes and the number of stress range cycles corresponding to each first stress cycle amplitude), where k = 1, 2, 3…, N. Based on the three-point rain flow counting method and the Miner-Palmgren damage accumulation theory, the fatigue damage during the monitoring time of the k-th fatigue hotspot is:

[0157] ;

[0158] in, is the fatigue damage of the kth fatigue hotspot during the monitoring time; is the number of stress range cycles after rainflow counting, that is, the number of stress range cycles corresponding to the amplitude of the i-th first stress cycle; is the average number of times the structure undergoes fatigue failure under the i-th constant amplitude stress (first stress cycle amplitude) based on the SN curve; m is the number of groups with different stress cycle amplitudes after rainflow counting, that is, the total number of first stress cycle amplitudes.

[0159] Step S52: Calculate fatigue damage caused by long-term environmental loads at each fatigue hotspot using a three-point rainflow counting method and a time-domain finite element method based on a preset stress-life curve, multiple preset floating platform loading conditions, and multiple short-term sea conditions corresponding to each local structure;

[0160] Furthermore, step S52 may include the following sub-steps S521-S524:

[0161] Step S521: using a time-domain finite element method to determine second stress time history data of multiple fatigue hot spots according to multiple short-term sea conditions and multiple preset floating platform loading conditions corresponding to each local structure;

[0162] Step S522: Analyze the second stress time history data of multiple fatigue hot spots using a three-point rainflow counting method to determine multiple second stress cycle amplitudes and the number of stress range cycles corresponding to each second stress cycle amplitude;

[0163] Step S523: determining the number of stress cycles required for fatigue failure of the structure corresponding to each second stress cycle amplitude based on the preset stress-life curve and the plurality of second stress cycle amplitudes;

[0164] Step S524: Calculate the fatigue damage caused by the long-term environmental load of each fatigue hotspot according to the number of stress cycles and the number of stress range cycles required for the structure to undergo fatigue failure corresponding to each second stress cycle amplitude.

[0165] It is important to note that to accurately predict the remaining fatigue life of an offshore platform, after determining the fatigue damage during the monitoring period, the fatigue damage during the platform's service life should also be calculated. This fatigue damage, which occurs before the platform is monitored, is composed of fatigue damage caused by both long-term and extreme environmental loads. Long-term environmental loads refer to the statistical characteristics of the environmental conditions experienced by the offshore platform throughout its design life. These loads reflect the combined impact of various environmental factors that the platform may encounter throughout its service life.

[0166] Furthermore, the frequency domain fatigue spectrum analysis method is used to calculate the fatigue damage caused by the long-term environmental load of the k-th fatigue hotspot based on the wave scatter diagram and SN curve of the sea area where the floating platform is in service, where k = 1, 2, 3…, N. Specifically, the time domain finite element calculation method is first used to input short-term sea conditions and loading conditions, that is, multiple short-term sea conditions corresponding to each local structure and multiple preset floating platform loading conditions, so as to obtain the fatigue hotspot stress response time history data (that is, the second stress time history data of multiple fatigue hotspots). Then, the three-point rain flow counting method is used to process the data to obtain a series of stress amplitudes and corresponding stress cycle numbers, that is, multiple second stress cycle amplitudes and the number of stress range cycles corresponding to each second stress cycle amplitude, so as to calculate the fatigue damage caused by the long-term environmental load of each fatigue hotspot. The specific process is as follows:

[0167] ;

[0168] in, is the fatigue damage caused by the long-term environmental load of the kth fatigue hotspot; The number of years in service before stress monitoring of floating platforms; is the number of different stress amplitudes under long-term environmental loads within one year, that is, the total number of the second stress cycle amplitudes; is the number of stress cycles of the i-th stress amplitude under long-term environmental load within one year, that is, the number of stress range cycles corresponding to the second stress cycle amplitude; It is the number of stress cycles required for the structure to undergo fatigue failure under the action of this stress amplitude, that is, the number of stress cycles required for the structure to undergo fatigue failure under the action of the i-th second stress cycle amplitude.

[0169] Step S53: using a three-point rainflow counting method and a time-domain finite element method, the fatigue damage of each fatigue hotspot under extreme working conditions is calculated according to the extreme environmental load conditions corresponding to the floating platform, multiple preset floating platform loading conditions, and preset stress-life curves;

[0170] Extreme environmental loads refer to the most severe environmental conditions an offshore platform may encounter during its service life, such as wave conditions that occur once every 50 years. Extreme environmental loads have a short duration but large stress amplitudes, so the fatigue damage they cause cannot be ignored.

[0171] Specifically, step S53 may include the following sub-steps S531-S534:

[0172] Step S531: using a time-domain finite element method to determine the third stress time history data of multiple fatigue hotspots according to the extreme environmental load conditions corresponding to the floating platform and multiple preset floating platform loading conditions;

[0173] Step S532: Analyze the third stress time history data of multiple fatigue hot spots using a three-point rainflow counting method to determine multiple third stress cycle amplitudes and the number of stress range cycles corresponding to each third stress cycle amplitude;

[0174] Step S533: determining the number of stress cycles required for fatigue failure of the structure corresponding to each third stress cycle amplitude based on the preset stress-life curve and the plurality of third stress cycle amplitudes;

[0175] Step S534: Calculate the fatigue damage of each fatigue hotspot under extreme working conditions according to the number of stress cycles and the number of stress range cycles required for fatigue failure of the structure corresponding to each third stress cycle amplitude.

[0176] It should be noted that the extreme loads (extreme environmental load conditions) encountered by the floating platform during its service life are counted, and the extreme environmental load conditions and platform loading conditions are input into the finite element calculation program. That is, the time-domain finite element method is used to obtain the stress response of the floating platform under extreme conditions (i.e., the third stress time history data of multiple fatigue hotspots) based on the extreme environmental load conditions corresponding to the floating platform and multiple preset floating platform loading conditions. The fatigue damage of the kth fatigue hotspot under extreme conditions (i.e., the fatigue damage of the fatigue hotspot under extreme conditions) is calculated using the three-point rain flow counting method and the Miner-Palmgren damage accumulation theory, where k = 1, 2, 3…, N.

[0177] ;

[0178] in, is the fatigue damage of the kth fatigue hotspot under extreme working conditions; is the number of cycles of the i-th stress amplitude caused by the j-th extreme working condition (extreme environmental load condition), that is, the number of cycles of the stress range corresponding to the i-th third stress cycle amplitude caused by the j-th extreme working condition; is the number of stress cycles required for fatigue failure to occur under the action of the i-th stress amplitude caused by the j-th extreme working condition, that is, the number of stress cycles required for fatigue failure of the structure under the action of the i-th third stress cycle amplitude caused by the j-th extreme working condition; is the number of different stress amplitudes under extreme working conditions, that is, the total number of third stress cycle amplitudes; m is the number of extreme working conditions encountered by the floating platform during its service life.

[0179] Step S54: summing up the fatigue damage during the monitoring period, the fatigue damage caused by long-term environmental loads, and the fatigue damage under extreme working conditions of each fatigue hotspot to determine the cumulative fatigue damage of each fatigue hotspot;

[0180] It should be noted that, based on the above foundation, the fatigue damage of the kth fatigue hotspot during the service period for:

[0181] ;

[0182] Furthermore, combined with the fatigue damage during the monitoring time, the cumulative fatigue damage of the kth fatigue hotspot for:

[0183] ;

[0184] Step S55: Calculate the annual fatigue damage and the remaining fatigue life of each fatigue hotspot based on the accumulated fatigue damage of each fatigue hotspot.

[0185] It should be noted that, assuming that the fatigue damage caused by the wave in the kth fatigue hotspot is the same each year, the fatigue damage in the kth fatigue hotspot each year is for:

[0186] ;

[0187] Wherein, T is the service life of the floating platform before stress monitoring, in years; t is the stress monitoring time of the floating platform, in years.

[0188] Furthermore, the remaining fatigue life of the kth fatigue hotspot is (Unit is year) is:

[0189] ;

[0190] For example, the cumulative fatigue damage of fatigue hotspots is composed of fatigue damage during the monitoring period, fatigue damage caused by long-term environmental loads during the service period, and fatigue damage caused by extreme environmental loads during the service period. Rainflow counting is performed on the stress response prediction data of the fatigue hotspot during the monitoring period, and the fatigue damage during this period is calculated in combination with the SN curve. The DNV-1 SN curve is used, and the curve parameters are shown in Table 3. The fatigue damage degree during the monitoring period is calculated to be 1.37×10 -16 .

[0191] Table 3 DNV-1 SN curve parameters

[0192]

[0193] Furthermore, using the South China Sea wave scatter diagram and JONSWAP wave spectrum, we assumed a 10-year service life for the floating platform, with equal probability for each wave direction. The parameters for the South China Sea wave scatter diagram are shown in Table 4. Furthermore, the platform was subjected to a 50-year extreme operating condition, with a significant wave height of 12.6 m, a significant period of 24.1 seconds, and a duration of three days, with equal probability for each wave direction. As shown in Table 5, the fatigue damage from long-term and extreme environmental loads during the service life was 0.43115, respectively. The fatigue damage degree per year at the fatigue hotspot was 0.043, and the remaining fatigue life was 12.77 years.

[0194] Table 4 Parameters of the South China Sea wave scatter diagram

[0195]

[0196] Table 5 Fatigue damage and remaining fatigue life of fatigue hot spots

[0197]

[0198] For a comparison of technical effectiveness, existing technologies can be used as a reference. To assess fatigue damage on floating platforms, commonly used fatigue monitoring techniques deploy a limited number of sensors to monitor stress responses under different sea conditions and perform linear extrapolation to obtain hotspot stress data. Fatigue analysis is then performed using the rainflow counting method in conjunction with the SN curve. However, floating platforms are large in size and have numerous fatigue hotspots. High deployment costs and limited physical space mean that some fatigue hotspots are not suitable for direct stress monitoring. Existing fatigue monitoring technologies are costly and have limited coverage. Floating platforms are large in size and have numerous fatigue hotspots. Furthermore, due to structural constraints, most areas are not suitable for sensor deployment, making this method incapable of comprehensively assessing fatigue damage on floating platforms. Furthermore, existing fatigue damage assessment methods that combine stress monitoring and finite element calculations have limited accuracy, leaving room for improvement. A floating platform fatigue damage assessment method based on the overall stress response combines stress monitoring and finite element calculations to establish two databases: one that uses the stress responses of the overall stress measurement points as input and wave parameters as output, and the other that uses wave parameters as input and fatigue damage rates at fatigue hotspots as output. This method calculates the cumulative fatigue damage value of fatigue hotspots based on the overall stress response monitoring data and database. However, this method calculates the fatigue damage of fatigue hotspots based on the fatigue spectrum analysis method, and therefore cannot take into account factors such as nonlinearity and low-frequency fatigue stress. Therefore, its accuracy is still insufficient compared to the time-domain fatigue calculation method. In addition, the existing floating platform fatigue damage assessment method does not take into account the cumulative fatigue damage during the service life, let alone the fatigue damage caused by extreme working conditions during the service life, and cannot accurately assess the remaining fatigue life of the floating platform. With global warming and the greenhouse effect, extreme working conditions such as typhoons and tsunamis are becoming more frequent, and the probability of extreme working conditions occurring has increased significantly. Failure to consider the fatigue damage caused by them will result in an overly large predicted fatigue life and insufficient safety margin.

[0199] Therefore, traditional stress monitoring methods are difficult to comprehensively and accurately assess the fatigue damage state of the structure. There is an urgent need for a low-cost and accurate floating platform fatigue damage assessment method that can solve the limited coverage of traditional fatigue monitoring.

[0200] In response to the above problems, the present invention provides a method for fatigue damage assessment of floating platforms, aiming to overcome the high cost and limited application of fatigue monitoring methods, while improving the accuracy of remaining fatigue life assessment. Figure 101) Machine Learning-Based Fatigue Hotspot Stress Inversion: Utilizing machine learning techniques, fatigue damage at fatigue hotspots is predicted based on the stress history of a limited number of monitoring points, thereby expanding the coverage of fatigue monitoring. A stress history database for monitoring points and fatigue hotspots is established through time-domain finite element calculations. Furthermore, utilizing machine learning algorithms, a prediction model is established with the stress history of monitoring points as input and the stress history of fatigue hotspots as output. Combined with the rainflow counting method and selected SN curves, the cumulative fatigue damage at fatigue hotspots during the stress monitoring period is calculated, achieving a comprehensive assessment of fatigue hotspots. 2) Improving Fatigue Damage Assessment Accuracy: Utilizing time-domain finite element calculations, the accuracy of fatigue assessment is significantly improved by fully considering factors such as nonlinearity and low-frequency fatigue stresses. Compared with traditional frequency-domain fatigue spectrum analysis methods, time-domain finite element calculations can more accurately capture complex stress response characteristics, thereby providing higher reliability for fatigue damage assessment. 3) Accurate Prediction of Remaining Fatigue Life: For fatigue damage within the service life of floating platforms, the effects of long-term and extreme environmental loads are considered to accurately predict their remaining fatigue life. By categorizing fatigue damage over the lifetime of a floating platform into two parts—long-term environmental loads and extreme environmental loads—and analyzing each separately, combined with fatigue damage predicted during the stress monitoring period using machine learning, a comprehensive assessment of the platform's cumulative fatigue damage in its current state is made. This further allows for prediction of its remaining fatigue life, providing a scientific basis for safe operation and maintenance decisions for the floating platform.

[0201] In summary, the present invention proposes an intelligent prediction model for non-measurement point stress (a fatigue hotspot stress response prediction model) for sensor-equipped offshore platform structures. This method achieves spatial correlation mapping of the structural stress field by constructing a multi-layer perceptual neural network, and achieves non-measurement point stress prediction by deeply mining the nonlinear correlation characteristics between measurement points. This effectively solves the technical problem of insufficient prediction accuracy of traditional interpolation methods in complex marine engineering environments. A dual-hidden layer network architecture is designed based on the structural characteristics of offshore platforms, effectively controlling model complexity while ensuring prediction accuracy. At the same time, the present invention combines stress monitoring with time-domain finite element calculation methods, comprehensively considering factors such as nonlinearity and low-frequency fatigue stress, and improving the accuracy of the traditional combined stress monitoring and frequency-domain finite element calculation method. In addition, the present invention also considers fatigue damage caused by long-term and extreme environmental loads during the service life of the floating platform, overcoming the shortcomings of methods that only consider fatigue damage caused by long-term environmental loads on the floating platform, and accurately predicting the remaining fatigue life of the floating platform.

[0202] In an embodiment of the present invention, the present invention provides a method for assessing fatigue damage of a floating platform. First, a wave scatter diagram of the working sea area of the floating platform is obtained, and a floating platform overall finite element model corresponding to the floating platform is established; then, based on the wave scatter diagram and a plurality of preset wave spectra, the long-term sea conditions in the wave scatter diagram are discretized to determine a plurality of short-term sea conditions corresponding to each local structure in the overall finite element model of the floating platform; a preset fatigue hotspot analysis method is used to perform fatigue hotspot analysis on the overall finite element model of the floating platform according to a plurality of preset regular wave sea conditions, wave scatter diagrams, a plurality of preset floating platform loading conditions and a plurality of short-term sea conditions corresponding to each local structure, to determine a plurality of fatigue hotspots; a time domain finite element method and a machine learning algorithm are used to analyze the fatigue hotspots, each preset wave condition and the short-term sea conditions corresponding to each local structure. A fatigue hotspot stress response prediction model is constructed by combining the floating platform loading conditions, various preset wave spectra, and multiple stress monitoring points of the floating platform. Finally, based on the three-point rainflow counting method and the time-domain finite element method, the fatigue hotspot stress response prediction model is used to generate fatigue damage results for each fatigue hotspot according to each stress monitoring point, the extreme environmental load conditions corresponding to the floating platform, multiple preset floating platform loading conditions, multiple short-term sea conditions corresponding to each local structure, and preset stress-life curves. Based on the above scheme, a fatigue hotspot stress response prediction model is established using a machine learning algorithm. Combined with the three-point rainflow counting method and the preset stress-life curve, the fatigue damage results for each fatigue hotspot in the floating platform are generated, thereby achieving a comprehensive assessment of the fatigue damage of the floating platform.

[0203] See also Figure 11 , Figure 11 A structural block diagram of a floating platform fatigue damage assessment device provided in an embodiment of the present invention.

[0204] The present invention provides a floating platform fatigue damage assessment device, comprising:

[0205] An acquisition module 1101 is used to obtain a wave scatter diagram of the working sea area of the floating platform and establish an overall finite element model of the floating platform corresponding to the floating platform;

[0206] A determination module 1102 is configured to discretize the long-term sea conditions in the wave scatter diagram based on the wave scatter diagram and a plurality of preset wave spectra, and determine a plurality of short-term sea conditions corresponding to each local structure in the overall finite element model of the floating platform;

[0207] An analysis module 1103 is configured to perform fatigue hotspot analysis on the entire finite element model of the floating platform using a preset fatigue hotspot analysis method based on multiple preset regular wave sea conditions, wave scatter diagrams, multiple preset floating platform loading conditions, and multiple short-term sea conditions corresponding to each local structure, thereby determining multiple fatigue hotspots.

[0208] A construction module 1104 is configured to construct a fatigue hotspot stress response prediction model using a time-domain finite element method and a machine learning algorithm based on each fatigue hotspot, each preset floating platform loading condition, each preset wave spectrum, and multiple stress monitoring points of the floating platform;

[0209] Evaluation module 1105 is used to generate fatigue damage results for each fatigue hotspot based on the three-point rainflow counting method and the time-domain finite element method, using a fatigue hotspot stress response prediction model according to each stress monitoring point, the extreme environmental load conditions corresponding to the floating platform, multiple preset floating platform loading conditions, multiple short-term sea conditions corresponding to each local structure, and preset stress-life curves.

[0210] Furthermore, the preset fatigue hotspot analysis method includes the classification rules for offshore mobile platforms, stress response under hazardous conditions, and fatigue spectrum analysis method; the determination module 1102 includes:

[0211] The first submodule is used to analyze the overall finite element model of the floating platform based on the offshore mobile platform classification rules and identify multiple fatigue hotspots;

[0212] The second submodule is used to analyze the overall finite element model of the floating platform based on the stress response under hazardous conditions and identify multiple fatigue hotspots;

[0213] The third submodule is used to analyze the overall finite element model of the floating platform using a fatigue spectrum analysis method based on multiple preset regular wave conditions, wave scatter diagrams, multiple preset floating platform loading conditions, and multiple short-term sea conditions corresponding to each local structure, to identify multiple fatigue hotspots.

[0214] Furthermore, the third submodule is specifically used to:

[0215] Constructing stress transfer functions for each local structure of the overall finite element model of the floating platform under various preset regular wave sea conditions and various preset floating platform loading conditions, and generating multiple stress transfer functions corresponding to each local structure;

[0216] Based on multiple short-term sea conditions corresponding to each local structure, the wave spectrum of each local structure under each short-term sea condition is determined;

[0217] Calculate the stress energy spectrum density of each local structure under each short-term sea condition based on multiple stress transfer functions corresponding to each local structure and the wave spectrum of each local structure under each short-term sea condition;

[0218] Based on the wave scatter diagram, the joint probability of significant wave height and zero-crossing period for each short-term sea state is determined;

[0219] The total fatigue life of each local structure is calculated based on the stress energy spectrum density of each local structure under each short-term sea state, the significant wave height of each short-term sea state, and the joint probability of the zero-crossing period.

[0220] The total fatigue life of each local structure is sorted in ascending order, and the local structure corresponding to the total fatigue life of the first preset number of digits is selected as the fatigue hotspot.

[0221] Furthermore, the construction module 1104 is specifically configured to:

[0222] The time-domain finite element method is used to output simulated stress response time history data of multiple stress monitoring points and simulated stress response time history data of multiple fatigue hotspots based on each preset floating platform loading condition, wave characteristic parameters corresponding to each preset wave spectrum, multiple stress monitoring points and multiple fatigue hotspots.

[0223] Based on the machine learning algorithm, a fatigue hotspot stress response prediction model is constructed, which takes the simulated stress response time history data of multiple stress monitoring points as input and the simulated stress response time history data of multiple fatigue hotspots as output.

[0224] Furthermore, the fatigue damage results include fatigue damage occurring each year and remaining fatigue life; the evaluation module 1105 includes:

[0225] The fourth submodule is used to calculate the fatigue damage of each fatigue hotspot within the monitoring time based on the monitoring stress response time history data corresponding to each stress monitoring point and the preset stress-life curve based on the three-point rain flow counting method and the fatigue hotspot stress response prediction model;

[0226] The fifth submodule is used to calculate the fatigue damage caused by long-term environmental loads at each fatigue hotspot using the three-point rainflow counting method and the time-domain finite element method based on preset stress-life curves, multiple preset floating platform loading conditions, and multiple short-term sea conditions corresponding to each local structure;

[0227] The sixth submodule is used to calculate the fatigue damage of each fatigue hotspot under extreme conditions using the three-point rainflow counting method and the time-domain finite element method based on the extreme environmental load conditions corresponding to the floating platform, multiple preset floating platform loading conditions, and preset stress-life curves;

[0228] The seventh submodule is used to sum up the fatigue damage during the monitoring period of each fatigue hotspot, the fatigue damage caused by long-term environmental loads, and the fatigue damage under extreme working conditions to determine the cumulative fatigue damage of each fatigue hotspot;

[0229] The eighth submodule is used to calculate the annual fatigue damage and remaining fatigue life of each fatigue hotspot based on the cumulative fatigue damage of each fatigue hotspot.

[0230] Furthermore, the fourth submodule is specifically configured to:

[0231] The monitored stress response time history data corresponding to each stress monitoring point is used as the input of the fatigue hotspot stress response prediction model, and the first stress response time history data of multiple fatigue hotspots is output;

[0232] The three-point rainflow counting method is used to analyze the first stress time history data of multiple fatigue hot spots to determine multiple first stress cycle amplitudes and the number of stress range cycles corresponding to each first stress cycle amplitude;

[0233] Determining, based on a preset stress-life curve and a plurality of first stress cycle amplitudes, an average number of fatigue failures of the structure corresponding to each first stress cycle amplitude;

[0234] The fatigue damage of each fatigue hotspot within the monitoring time is calculated based on the average number of fatigue failures of the structure corresponding to each first stress cycle amplitude and the number of stress range cycles.

[0235] Furthermore, the fifth submodule is specifically configured to:

[0236] The time-domain finite element method is used to determine the secondary stress time history data of multiple fatigue hotspots based on multiple short-term sea conditions and multiple pre-set floating platform loading conditions corresponding to each local structure;

[0237] The three-point rainflow counting method is used to analyze the second stress time history data of multiple fatigue hot spots to determine multiple second stress cycle amplitudes and the number of stress range cycles corresponding to each second stress cycle amplitude;

[0238] Determining the number of stress cycles required for fatigue failure of the structure corresponding to each second stress cycle amplitude based on a preset stress-life curve and a plurality of second stress cycle amplitudes;

[0239] According to the number of stress cycles and stress range cycles required for the structure to undergo fatigue failure corresponding to each second stress cycle amplitude, the fatigue damage caused by the long-term environmental load of each fatigue hotspot is calculated.

[0240] Furthermore, the sixth submodule is specifically configured to:

[0241] The time-domain finite element method is used to determine the third stress time history data of multiple fatigue hotspots based on the extreme environmental load conditions corresponding to the floating platform and multiple preset floating platform loading conditions;

[0242] The three-point rainflow counting method is used to analyze the third stress time history data of multiple fatigue hot spots to determine multiple third stress cycle amplitudes and the number of stress range cycles corresponding to each third stress cycle amplitude;

[0243] Determining the number of stress cycles required for fatigue failure of the structure corresponding to each third stress cycle amplitude based on a preset stress-life curve and a plurality of third stress cycle amplitudes;

[0244] According to the number of stress cycles and stress range cycles required for the structure to undergo fatigue failure corresponding to each third stress cycle amplitude, the fatigue damage of each fatigue hotspot under extreme working conditions is calculated.

[0245] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described devices, modules and sub-modules can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0246] An embodiment of the present invention further provides a computer device comprising a memory and a processor, wherein a computer program is stored in the memory; when the computer program is executed by the processor, the processor executes the steps of the floating platform fatigue damage assessment method as described in any of the above embodiments.

[0247] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions described in the above embodiments can still be modified, or some of the technical features thereof can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for fatigue damage assessment of a floating platform, characterized in that: include: Obtaining a wave scatter diagram of the working sea area of the floating platform and establishing an overall finite element model of the floating platform corresponding to the floating platform; Based on the wave scatter diagram and a plurality of preset wave spectra, discretizing the long-term sea conditions in the wave scatter diagram, and determining a plurality of short-term sea conditions corresponding to each local structure in the overall finite element model of the floating platform; performing a fatigue hotspot analysis on the entire finite element model of the floating platform using a preset fatigue hotspot analysis method based on a plurality of preset regular wave sea conditions, the wave scatter diagram, a plurality of preset floating platform loading conditions, and a plurality of short-term sea conditions corresponding to each of the local structures, to determine a plurality of fatigue hotspots; A fatigue hotspot stress response prediction model is constructed using a time-domain finite element method and a machine learning algorithm based on each fatigue hotspot, each preset floating platform loading condition, each preset wave spectrum, and multiple stress monitoring points of the floating platform; Based on the three-point rainflow counting method and the time-domain finite element method, the fatigue hotspot stress response prediction model is used to generate fatigue damage results for each fatigue hotspot according to each stress monitoring point, the extreme environmental load conditions corresponding to the floating platform, multiple preset floating platform loading conditions, multiple short-term sea conditions corresponding to each local structure, and preset stress-life curves.

2. The floating platform fatigue damage assessment method according to claim 1, characterized in that: The preset fatigue hotspot analysis method includes offshore mobile platform classification rules, stress response under hazardous conditions, and fatigue spectrum analysis methods; the preset fatigue hotspot analysis method is used to perform fatigue hotspot analysis on the entire finite element model of the floating platform based on multiple preset regular wave conditions, the wave scatter diagram, multiple preset floating platform loading conditions, and multiple short-term sea conditions corresponding to each of the local structures, to determine multiple fatigue hotspots, including: Based on the offshore mobile platform classification rules, analyzing the overall finite element model of the floating platform to identify multiple fatigue hotspots; Analyzing the overall finite element model of the floating platform based on the stress response under the hazardous working condition to identify multiple fatigue hotspots; The fatigue spectrum analysis method is used to analyze the overall finite element model of the floating platform according to the multiple preset regular wave sea conditions, the wave scatter diagram, multiple preset floating platform loading conditions and multiple short-term sea conditions corresponding to each of the local structures to determine multiple fatigue hotspots.

3. The floating platform fatigue damage assessment method according to claim 2, characterized in that: The fatigue spectrum analysis method is used to analyze the entire finite element model of the floating platform according to the plurality of preset regular wave sea conditions, the wave scatter diagram, a plurality of preset floating platform loading conditions, and a plurality of short-term sea conditions corresponding to each of the local structures to determine a plurality of fatigue hot spots, including: constructing stress transfer functions for each local structure of the overall finite element model of the floating platform under each of the preset regular wave sea conditions and each of the preset floating platform loading conditions, and generating a plurality of stress transfer functions corresponding to each of the local structures; determining, based on a plurality of short-term sea conditions corresponding to each of the local structures, a wave spectrum of each of the local structures under each of the short-term sea conditions; Calculating the stress energy spectrum density of each local structure under each of the short-term sea conditions based on a plurality of stress transfer functions corresponding to each of the local structures and a wave spectrum of each of the local structures under each of the short-term sea conditions; Determining the joint probability of significant wave height and zero-crossing period for each of the short-term sea conditions based on the wave scatter diagram; Calculating the total fatigue life of each of the local structures according to the stress energy spectrum density of each of the local structures under each of the short-term sea conditions, the significant wave height of each of the short-term sea conditions, and the joint probability of the zero-crossing period; The total fatigue life of each of the local structures is sorted in ascending order, and the local structure corresponding to the total fatigue life of the first preset number of digits is selected as the fatigue hotspot.

4. The floating platform fatigue damage assessment method according to claim 1, characterized in that: The time-domain finite element method and the machine learning algorithm are used to construct a fatigue hotspot stress response prediction model based on each fatigue hotspot, each preset floating platform loading condition, each preset wave spectrum, and multiple stress monitoring points of the floating platform, including: outputting simulated stress response time history data of the plurality of stress monitoring points and the simulated stress response time history data of the plurality of fatigue hotspots based on the preset floating platform loading conditions, the wave characteristic parameters corresponding to the preset wave spectra, the positions of the plurality of stress monitoring points and the plurality of fatigue hotspots using a time-domain finite element method; Based on a machine learning algorithm, a fatigue hotspot stress response prediction model is constructed, which takes the simulated stress response time history data of multiple stress monitoring points as input and takes the simulated stress response time history data of multiple fatigue hotspots as output.

5. The floating platform fatigue damage assessment method according to claim 1, characterized in that: The fatigue damage results include fatigue damage occurring each year and remaining fatigue life; the fatigue damage results of each fatigue hotspot are generated by using the fatigue hotspot stress response prediction model based on the three-point rainflow counting method and the time-domain finite element method according to each stress monitoring point, the extreme environmental load conditions corresponding to the floating platform, multiple preset floating platform loading conditions, multiple short-term sea conditions corresponding to each local structure, and preset stress-life curves, including: Based on the three-point rainflow counting method, the fatigue hotspot stress response prediction model is used to calculate the fatigue damage of each fatigue hotspot within the monitoring time according to the monitored stress response time history data corresponding to each stress monitoring point and the preset stress-life curve; Calculating fatigue damage caused by long-term environmental loads at each fatigue hotspot using the three-point rainflow counting method and the time-domain finite element method based on the preset stress-life curve, multiple preset floating platform loading conditions, and multiple short-term sea conditions corresponding to each local structure; The three-point rainflow counting method and the time-domain finite element method are used to calculate the fatigue damage of each fatigue hotspot under extreme working conditions according to the extreme environmental load conditions corresponding to the floating platform, a plurality of preset floating platform loading conditions, and the preset stress-life curve; The fatigue damage of each fatigue hotspot during the monitoring period, the fatigue damage caused by long-term environmental loads, and the fatigue damage under extreme working conditions are respectively added to determine the cumulative fatigue damage of each fatigue hotspot; Based on the accumulated fatigue damage of each fatigue hotspot, the fatigue damage occurring each year and the remaining fatigue life of each fatigue hotspot are calculated.

6. The floating platform fatigue damage assessment method according to claim 5, characterized in that: The three-point rainflow counting method is based on the fatigue hotspot stress response prediction model, and the fatigue damage within the monitoring time of each fatigue hotspot is calculated according to the monitored stress response time history data corresponding to each stress monitoring point and the preset stress-life curve, including: using the monitored stress response time history data corresponding to each of the stress monitoring points as input to the fatigue hotspot stress response prediction model, and outputting first stress response time history data of a plurality of fatigue hotspots; Analyzing the first stress time history data of the plurality of fatigue hot spots using the three-point rainflow counting method to determine a plurality of first stress cycle amplitudes and the number of stress range cycles corresponding to each of the first stress cycle amplitudes; determining, based on the preset stress-life curve and a plurality of the first stress cycle amplitudes, an average number of times fatigue failure of the structure occurs corresponding to each of the first stress cycle amplitudes; The fatigue damage of each fatigue hotspot within the monitoring time is calculated according to the average number of times fatigue failure of the structure occurs and the number of stress range cycles corresponding to each of the first stress cycle amplitudes.

7. The floating platform fatigue damage assessment method according to claim 5, characterized in that: The method of using the three-point rainflow counting method and the time-domain finite element method to calculate the fatigue damage caused by the long-term environmental load of each fatigue hotspot according to the preset stress-life curve, multiple preset floating platform loading conditions, and multiple short-term sea conditions corresponding to each local structure includes: Determining the second stress time history data of the plurality of fatigue hot spots using the time domain finite element method according to the plurality of short-term sea conditions corresponding to the respective local structures and the plurality of preset floating platform loading conditions; Analyzing the second stress time history data of the plurality of fatigue hot spots using the three-point rainflow counting method to determine a plurality of second stress cycle amplitudes and the number of stress range cycles corresponding to each of the second stress cycle amplitudes; determining, based on the preset stress-life curve and a plurality of second stress cycle amplitudes, the number of stress cycles required for fatigue failure of the structure corresponding to each second stress cycle amplitude; The fatigue damage caused by the long-term environmental load of each fatigue hotspot is calculated based on the number of stress cycles and the number of stress range cycles required for the structure to undergo fatigue failure corresponding to each second stress cycle amplitude.

8. The floating platform fatigue damage assessment method according to claim 5, characterized in that: The method of using the three-point rainflow counting method and the time-domain finite element method to calculate the fatigue damage of each fatigue hotspot under extreme working conditions according to the extreme environmental load conditions corresponding to the floating platform, a plurality of preset floating platform loading conditions, and the preset stress-life curve includes: Determining the third stress time history data of the plurality of fatigue hot spots using the time domain finite element method according to the extreme environmental load conditions corresponding to the floating platform and the plurality of preset floating platform loading conditions; Analyzing the third stress time history data of the plurality of fatigue hot spots using the three-point rainflow counting method to determine a plurality of third stress cycle amplitudes and the number of stress range cycles corresponding to each third stress cycle amplitude; determining, based on the preset stress-life curve and the plurality of third stress cycle amplitudes, the number of stress cycles required for fatigue failure of the structure corresponding to each of the third stress cycle amplitudes; The fatigue damage of each fatigue hotspot under extreme working conditions is calculated based on the number of stress cycles and the number of stress range cycles required for fatigue failure of the structure corresponding to each of the third stress cycle amplitudes.

9. A floating platform fatigue damage assessment device, characterized in that: include: An acquisition module is used to obtain a wave scatter diagram of the working sea area of the floating platform and establish an overall finite element model of the floating platform corresponding to the floating platform; a determination module, configured to discretize the long-term sea conditions in the wave scatter diagram based on the wave scatter diagram and a plurality of preset wave spectra, and determine a plurality of short-term sea conditions corresponding to each local structure in the overall finite element model of the floating platform; an analysis module, configured to perform fatigue hotspot analysis on the entire finite element model of the floating platform using a preset fatigue hotspot analysis method according to a plurality of preset regular wave sea conditions, the wave scatter diagram, a plurality of preset floating platform loading conditions, and a plurality of short-term sea conditions corresponding to each of the local structures, to determine a plurality of fatigue hotspots; a construction module for constructing a fatigue hotspot stress response prediction model based on each fatigue hotspot, each preset floating platform loading condition, each preset wave spectrum, and a plurality of stress monitoring points of the floating platform using a time-domain finite element method and a machine learning algorithm; An evaluation module is configured to generate fatigue damage results for each fatigue hotspot based on a three-point rainflow counting method and the time-domain finite element method, using the fatigue hotspot stress response prediction model according to each stress monitoring point, the extreme environmental load conditions corresponding to the floating platform, multiple preset floating platform loading conditions, multiple short-term sea conditions corresponding to each local structure, and preset stress-life curves.

10. A computer device, characterized in that: The method comprises a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the processor executes the steps of the floating platform fatigue damage assessment method according to any one of claims 1 to 8.

Images