Method and device for analyzing fatigue damage of floating wind turbine floating body structure

Abstract

The application provides a floating fan floating body structure fatigue damage analysis method and system, relates to the technical field of fan floating body structure fatigue damage analysis method, and comprises the following steps: obtaining total load response at a target analysis section; calculating an RMS correction factor, correcting the total load response of all analysis sections of each working condition, and selecting an optimal stress amplitude probability density function; recalculating a frequency domain short-term fatigue damage coefficient for all working conditions, obtaining an optimal short-term fatigue damage coefficient, calculating a long-term fatigue damage coefficient, and evaluating fatigue life. The frequency domain fatigue analysis is introduced, the damage is directly calculated by using the stress power spectrum and the probability density function, the processing speed is greatly improved, the stress amplitude probability density function most suitable for the current working condition and the section is adaptively selected by comparing the frequency domain damage and the time domain damage, and the calculation accuracy is improved. The RMS correction factor is calculated, the linear superposition result is corrected, and the high efficiency of the wind wave decoupling is retained.

Description

Technical Field

[0001] This invention relates to the technical field of fatigue damage analysis methods for floating structures of wind turbines, specifically to a fatigue damage analysis method and system for floating structures of floating wind turbines. Background Technology

[0002] As core equipment for deep-sea wind energy development, offshore floating wind turbines endure the combined effects of complex environmental loads such as wind, waves, and currents over long periods, posing a severe risk of fatigue damage to their floating structures. Currently, the fully coupled time-domain analysis method is commonly used in engineering for fatigue assessment. This method performs complete wind-wave-structure coupled simulations for each type of marine environmental condition, directly obtaining stress time histories before performing rainflow counting and damage calculations. However, due to the diversity of wind and wave parameter combinations and the complexity of coupling effects, fully coupled time-domain analysis requires extensive simulation calculations, typically taking weeks or even months to complete a full long-term fatigue assessment, severely limiting the efficiency of iterative optimization in engineering design. Furthermore, traditional methods demand extremely high computational resources and are difficult to apply flexibly to different types of floating structures while maintaining accuracy.

[0003] In the existing technology, document CN120030947A proposes a method to decouple wind and waves, calculate the wind load response at the structural cross-section of the floating structure under different wind conditions and the wave load response at the structural cross-section of the floating structure under different wave actions, and then perform stress superposition in the post-processing stage. However, this method does not use recursive iterative cutting to construct an equal probability set of working conditions, discretizing the continuous wind and wave parameter space into a finite number of representative working conditions; it does not use a wind and wave decoupling strategy to calculate the wind and wave load responses separately and perform linear superposition, but only needs to perform fully coupled simulation on a small number of representative working conditions to calculate the RMS correction factor, that is, to effectively correct the linear superposition results; it also does not introduce a frequency domain fatigue analysis method to adaptively select the optimal stress amplitude probability density function by comparing the damage results in the frequency domain and time domain, and its damage assessment accuracy is not high. Therefore, there is an urgent need for a fatigue damage analysis method and system for floating wind turbine floating structures.

[0004] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for fatigue damage analysis of floating structure of floating wind turbines, so as to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: A method for fatigue damage analysis of the floating body structure of a floating wind turbine, including the following steps: S1: Construct a finite element model of the floating structure to be analyzed, and label the cross-sections between adjacent structural segments within the floating structure as analysis cross-sections, and determine the target analysis cross-section from the analysis cross-sections; S2: Obtain several types of wind and wave parameters from the historical data of the floating structure deployment area, and use recursive iterative cutting to cut the wind and wave parameters to construct a set of working conditions. Using each type of working condition in the set of working conditions as a constraint, perform time-domain simulation on the finite element model to obtain the total load response at the target analysis section. S3: Calculate the probability of each working condition in the set of working conditions and determine the representative set of working conditions. Based on the representative set of working conditions, perform a fully coupled time-domain simulation of the entire floating structure of the floating wind turbine containing the target analysis section. Extract the fully coupled load response at the target analysis section from the simulation results as the verification working condition parameter. Calculate the RMS correction factor based on the verification working condition parameter and correct the total load response of the target analysis section under each working condition. S4: Based on the geometric characteristics of the target analysis section, the Euler-Bernoulli beam method is used to convert the corrected total load response into local stress. Fourier transform analysis is performed on the local stress to calculate the short-term fatigue damage coefficient in the frequency domain. Based on the local stress and the rainflow count in the time domain, and combined with the SN curve, the short-term fatigue damage coefficient in the time domain is calculated. The results of the short-term fatigue damage coefficient in the frequency domain and the short-term fatigue damage coefficient in the time domain are compared to select the optimal stress amplitude probability density function for the target analysis section under each working condition. S5: Based on the optimal stress amplitude probability density function, recalculate the frequency domain short-term fatigue damage coefficient for all working conditions to obtain the optimal short-term fatigue damage coefficient: Integrate the optimal short-term fatigue damage under all working conditions, and combine it with the probability under each working condition to calculate the long-term fatigue damage coefficient and evaluate the fatigue life.

[0007] Furthermore, a recursive iterative cutting method is used to construct the set of working conditions, specifically as follows: Several types of wind and wave parameters from the historical data of the floating structure deployment area are obtained, including average wind speed, significant wave height, spectral peak period, and wind and wave direction. These parameters are then labeled as a space of data points representing a four-dimensional parameter combination. After sorting all observations for each dimension, the midpoint between each adjacent observation point is sequentially considered as a candidate cutting point. The number of data points contained in the left and right subspaces after the cutting is calculated, and the cutting point that makes the number of points in the left and right subspaces closest to equal is selected as the optimal cutting point for that dimension. The optimal cutting results in the four dimensions are compared, and the dimension with the smallest difference in the number of data points in the left and right subspaces and its corresponding cutting point are selected for actual cutting, thus dividing the current space into two. Repeat the above process for each newly generated subspace, recursively cutting it until the number of subspaces obtained reaches the preset number. These final subspaces are the equally probable working condition intervals. The average value of each type of dimension parameter within each working condition interval is marked as the final parameter of the working condition interval, forming a discrete set of working conditions.

[0008] Furthermore, the total load response at the corresponding target analysis section is calculated, specifically as follows: Using each type of working condition in the set of working conditions as a constraint, a time-domain simulation is performed on the finite element model. The logic for the time-domain simulation is as follows: In the simulation, a wind field based on average wind speed and wave direction is applied. The wind load response data at the target analysis section under wind action is obtained from the simulation results. The wind load response data consists of an axial force time series and a surrounding... Bending moment time series of the shaft and its surroundings The bending moment time series of the shaft together constitute the structure; In the simulation, a wave field based on significant wave height, spectral peak period, and wave direction is applied. Based on the simulation results, wave load response data at the target analysis section under different wave actions are obtained. The wave load response data also consists of axial force time series and circumferential... Bending moment time series of the shaft and its surroundings The bending moment time series of the shaft is constituted; The wind load response data and wave load response data under each working condition are added together to obtain the total load response at the target analysis section under each working condition.

[0009] Further, the RMS correction factor is calculated as follows: The working conditions are sorted from high to low according to the probability weight of each working condition in the set of working conditions, and the working conditions with a cumulative probability of more than 80% are selected in turn to form a representative set of working conditions. Based on a representative set of operating conditions, a fully coupled time-domain simulation of the entire floating wind turbine system, including the target analysis section, is performed. The fully coupled load response at the target analysis section is extracted from the simulation results as verification operating condition parameters. Each verification operating condition parameter includes the fully coupled axial force time series and the load response around the target analysis section. Fully coupled bending moment time series of the shaft and around Fully coupled bending moment time series of the shaft; For each verification condition, the axial force time series under wind action and the axial force time series under wave action are first added together to obtain a linear superimposed axial force time series. The root mean square values ​​of the fully coupled axial force time series and the linear superimposed axial force time series are calculated respectively. The specific calculation method is as follows: the value at each moment in the time series is squared, all squared values ​​are integrated over the simulation time, the integration result is divided by the simulation time, and the square root of the quotient is taken to obtain the root mean square value of the fully coupled axial force. The root mean square value of the fully coupled axial force is divided by the root mean square value of the linearly superimposed axial force time series to obtain the axial force ratio for this verification condition. Using the same steps, the bending moments about the y-axis and z-axis are calculated separately to obtain the bending moment ratios about the y-axis and z-axis. The axial force ratios of all verification conditions are summed and divided by the total number of verification conditions to obtain the axial force RMS correction factor for the target analysis section. The bending moment ratios about the y-axis of all verification conditions are summed and divided by the total number of verification conditions to obtain the bending moment RMS correction factor about the y-axis. The bending moment ratios about the z-axis of all verification conditions are summed and divided by the total number of verification conditions to obtain the bending moment RMS correction factor about the z-axis.

[0010] Furthermore, the corrected total load response is obtained, specifically: First, the axial force time series of the wind field and wave field under this working condition are added together to obtain a linearly superimposed axial force time series. Then, the axial force RMS correction factor is multiplied by this linearly superimposed axial force time series to obtain the corrected axial force time series. For the bending moment about the y-axis, the y-axis bending moment time series of the wind field and wave field are added together to obtain a linearly superimposed y-axis bending moment time series. Then, the y-axis bending moment RMS correction factor is multiplied by this linearly superimposed y-axis bending moment time series to obtain the corrected y-axis bending moment time series. For the bending moment about the z-axis, the same steps are used, multiplying the z-axis bending moment RMS correction factor by the linearly superimposed z-axis bending moment time series to obtain the corrected z-axis bending moment time series. Finally, the corrected axial force time series, the corrected y-axis bending moment time series, and the corrected z-axis bending moment time series are combined to form the corrected total load response for this working condition.

[0011] Furthermore, the short-term fatigue damage coefficient in the frequency domain is calculated as follows: Divide each value in the corrected axial force time series of the target analysis section under each working condition by the cross-sectional area of ​​that section to obtain the stress time series generated by the axial force; take the absolute value of each value in the corrected bending moment time series of the target analysis section about the y-axis under each working condition and divide it by the section modulus of that section about the y-axis to obtain the stress time series generated by bending about the y-axis; take the absolute value of each value in the corrected bending moment time series of the target analysis section about the z-axis under each working condition and divide it by the section modulus of that section about the z-axis to obtain the stress time series generated by bending about the z-axis; add the values ​​of these three stress time series at corresponding times to obtain the local stress time series of the target analysis section under each working condition. The stress power spectrum is obtained by performing a Fourier transform on the local stress time series. The specific calculation method is as follows: For each frequency value, multiply each value in the local stress time series by an exponential function with the natural constant as the base and negative imaginary units multiplied by the frequency and the corresponding time. Summate all the products within the range of zero to the simulation duration to obtain the Fourier transform result at that frequency. Take the square of the modulus of the Fourier transform result and divide it by the simulation duration. When the simulation duration approaches infinity, this value is the stress power spectrum value at that frequency. For the zeroth spectral moment, the stress power spectrum is integrated over a frequency range from zero to infinity, i.e., the stress power spectrum values ​​corresponding to all frequencies are multiplied by the frequency interval and then summed. For the first spectral moment, each frequency value is multiplied by the stress power spectrum value at that frequency, and then integrated over the frequency range. For the second spectral moment, the square of the frequency is multiplied by the stress power spectrum value, and then integrated. For the fourth spectral moment, the fourth power of the frequency is multiplied by the stress power spectrum value, and then integrated, thus obtaining the zeroth, first, second, and fourth spectral moments, respectively. Based on the calculated spectral moments, four different stress amplitude probability density function models were compared, including the Rayleigh model, Dirlik model, Wirsching model, and Zhao-Baker model. For each probability density function model, short-term fatigue damage in the frequency domain was calculated. The specific calculation method was as follows: the stress range was used as the integration variable. First, the slope of the SN curve of the stress range was calculated to the power of the slope. Then, it was multiplied by the probability density of the stress range under the model. The product was integrated over the stress range from zero to infinity to obtain the integral result. The integral result was multiplied by the simulation duration and then divided by the material constant of the SN curve to obtain the short-term fatigue damage coefficient in the frequency domain under the model.

[0012] Furthermore, the optimal stress amplitude probability density function is selected as follows: Based on local stress and time-domain rainflow counting, and combined with SN curves, the short-term fatigue damage coefficient of the target analysis section under each working condition is calculated as follows: the peak and valley values ​​in the local stress time series are extracted sequentially and paired according to the rules of rainflow counting to identify complete stress cycles; for each identified stress cycle, the difference between its maximum and minimum values ​​is calculated to obtain the stress range of the cycle; all stress cycles are grouped according to the size of the stress range, and the number of cycles contained in each stress range is counted to obtain the value of each stress range and the number of times the stress range appears; The allowable number of cycles for each stress range is calculated based on the SN curve. The specific calculation method is as follows: divide the reference plate thickness by the actual plate thickness and calculate the ratio between the two; multiply the ratio by the thickness effect exponent to obtain the thickness correction factor; multiply each stress range by the thickness correction factor to obtain the corrected stress range; take the negative power of the slope of the SN curve for the corrected stress range and multiply it by the intercept of the SN curve to obtain the allowable number of cycles for that stress range. For each stress range, divide the number of times the stress range appears by the corresponding allowable number of cycles to obtain the fatigue damage caused by the stress range. Add up the fatigue damage of all groups to obtain the time-domain short-term fatigue damage coefficient of the target analysis section under each working condition. The short-term frequency domain fatigue damage results are compared with the short-term time domain fatigue damage coefficient results to calculate the relative error. The stress amplitude probability density function model that minimizes the relative error is selected as the optimal stress amplitude probability density function.

[0013] Further, fatigue life is assessed, specifically as follows: Based on the optimal stress amplitude probability density function, the frequency domain short-term fatigue damage coefficient is recalculated for all working conditions to obtain the optimal short-term fatigue damage coefficient. The stress range is used as the integration variable. First, the slope parameter of the SN curve of the stress range is calculated and then multiplied by the probability density of the stress range given by the optimal stress amplitude probability density function under this working condition. The product is integrated over the stress range from zero to infinity to obtain the integration result. The integration result is multiplied by the simulation duration and then divided by the SN curve constant to obtain the optimal short-term fatigue damage coefficient of the target analysis section under this working condition. The optimal short-term fatigue damage coefficients for all working conditions are summed, and the long-term fatigue damage coefficient is calculated by combining the probability of each working condition and the ratio of the design life to the simulation duration. Specifically, the optimal short-term fatigue damage coefficient for each working condition is multiplied by the probability of that working condition, and then multiplied by the quotient of the design life divided by the simulation duration to obtain the contribution of that working condition to long-term fatigue damage. The contributions of all working conditions are then summed to obtain the long-term fatigue damage coefficient of the target analysis section. The fatigue life of the target analysis section is assessed based on the long-term fatigue damage coefficient. The specific calculation method is as follows: divide the value by the long-term fatigue damage coefficient, and the quotient is the fatigue life of the target analysis section.

[0014] The present invention also provides a fatigue damage analysis system for the floating body structure of a floating wind turbine, the analysis system being used to perform the above-described analysis method, including: Data acquisition module: used to construct the finite element model of the floating structure to be analyzed, and to label the cross-sections between adjacent structural segments within the floating structure as analysis cross-sections, and to determine the target analysis cross-section from the analysis cross-sections; Total load calculation module: used to obtain several types of wind and wave parameters in the historical area of ​​the floating structure deployment area. The wind and wave parameters are cut by recursive iteration to construct a set of working conditions. Each type of working condition in the set of working conditions is used as a constraint condition to perform time-domain simulation on the finite element model to obtain the total load response at the target analysis section. Correction module: Calculates the probability of each working condition in the set of working conditions and determines the representative set of working conditions. Based on the representative set of working conditions, it performs a fully coupled time-domain simulation of the entire floating structure of the floating wind turbine containing the target analysis section. It extracts the fully coupled load response at the target analysis section from the simulation results as the verification working condition parameter. Based on the verification working condition parameter, it calculates the RMS correction factor and corrects the total load response of the target analysis section under each working condition. The Fourier transform module is used to convert the corrected total load response into local stress based on the geometric characteristics of the target analysis section using the Euler-Bernoulli beam method. Fourier transform analysis is performed on the local stress to calculate the short-term fatigue damage coefficient in the frequency domain. Based on the local stress and the rainflow count in the time domain, and combined with the SN curve, the short-term fatigue damage coefficient in the time domain is calculated. The results of the short-term fatigue damage coefficient in the frequency domain are compared with those in the time domain to select the optimal stress amplitude probability density function for the target analysis section under each working condition. Life assessment module: It is used to recalculate the frequency domain short-term fatigue damage coefficient for all working conditions based on the optimal stress amplitude probability density function to obtain the optimal short-term fatigue damage coefficient. It integrates the optimal short-term fatigue damage under all working conditions and combines it with the probability under each working condition to calculate the long-term fatigue damage coefficient and assess fatigue life.

[0015] Compared with the prior art, the beneficial effects of the present invention are: By employing recursive iterative cutting to construct an equiprobable set of working conditions, the continuous wind and wave parameter space is discretized into a finite number of representative working conditions, significantly reducing the number of working conditions required for subsequent simulations. Secondly, by using a wind-wave decoupling strategy to calculate the wind and wave load responses separately and then linearly superimpose them, only a small number of representative working conditions need to be fully coupled for simulation to calculate the RMS correction factor, which can effectively correct the linear superposition results, reducing the computational load while ensuring computational accuracy. Finally, a frequency domain fatigue analysis method is introduced, and the optimal stress amplitude probability density function is adaptively selected by comparing the damage results in the frequency domain and the time domain. This avoids the large amount of computation required by traditional time domain rainflow counting and ensures the accuracy of damage assessment. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the overall method flow of the present invention; Figure 2 A graph showing the relationship between the fatigue life of the target section and the long-term fatigue damage coefficient. Figure 3 This is a schematic diagram of the overall system structure of the present invention. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0018] It should be noted that, unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0019] Example: Please see Figures 1-2 The present invention provides a technical solution: A method for fatigue damage analysis of the floating body structure of a floating wind turbine, including the following steps: S1: Construct a finite element model of the floating structure to be analyzed, and label the cross-sections between adjacent structural segments within the floating structure as analysis cross-sections, and determine the target analysis cross-section from the analysis cross-sections; In the above process, the specific method for determining the target analysis section from the analysis section is as follows: identify areas with significant stress concentration, such as geometric abrupt changes and welding joints, namely the connection between the column and the cross brace, the connection between the column and the tower, around the opening, the end of the stiffening rib, and the intersection of components connected by welding processes, such as the intersection weld between the strut and the column, and the fillet weld between plates. These areas will have uneven stress distribution due to geometric discontinuity. Combining the classification society specifications and the engineering experience of similar floating wind turbines, we determine the key areas of concern and select 3 to 5 of the most representative sections as target analysis sections for detailed fatigue assessment.

[0020] S2: Obtain several types of wind and wave parameters from the historical data of the floating structure deployment area, and use recursive iterative cutting to cut the wind and wave parameters to construct a set of working conditions. Using each type of working condition in the set of working conditions as a constraint, perform time-domain simulation on the finite element model to obtain the total load response at the target analysis section. The set of working conditions is constructed by recursive iterative cutting, specifically as follows: Several types of wind and wave parameters from the historical data of the floating structure deployment area are obtained, including average wind speed, significant wave height, spectral peak period, and wind and wave direction. These parameters are then labeled as a space of data points representing a four-dimensional parameter combination. After sorting all observations for each dimension, the midpoint between each adjacent observation point is sequentially considered as a candidate cutting point. The number of data points contained in the left and right subspaces after the cutting is calculated, and the cutting point that makes the number of points in the left and right subspaces closest to equal is selected as the optimal cutting point for that dimension. The optimal cutting results in the four dimensions are compared, and the dimension with the smallest difference in the number of data points in the left and right subspaces and its corresponding cutting point are selected for actual cutting, thus dividing the current space into two. Repeat the above process for each newly generated subspace, recursively cutting it until the number of subspaces obtained reaches the preset number. These final subspaces are the equally probable working condition intervals. The average value of each type of dimension parameter within each working condition interval is marked as the final parameter of the working condition interval, forming a discrete set of working conditions.

[0021] In the above process, the continuous four-dimensional wind and wave parameter space is adaptively divided into several equally probable operating condition intervals. No preset parameter thresholds or grid boundaries are required; the division is based entirely on the actual distribution of historical observation data, ensuring that each operating condition interval contains approximately the same number of data points, thus giving each operating condition the same probability weight. This data-driven division method automatically focuses on high-probability areas with dense data, ensuring fine division of high-probability operating conditions while avoiding over-subdivision of low-probability areas, thereby minimizing the number of operating conditions while maintaining statistical representativeness. The recursive iterative division method essentially divides the space based on the actual distribution of historical observation data. By finding the cutting point in each dimension that makes the number of data points in the left and right subspaces approximately equal, it ensures that each final subspace contains approximately the same number of data points. Since each data point represents a set of actually observed wind and wave parameter combinations, the data point density directly reflects the probability density of that parameter combination occurring in the actual marine environment. Therefore, when each subspace contains an equal number of data points, the operating condition intervals represented by each subspace have statistically equal probability weights. The number of subspaces can be set according to the balance between computing resources and accuracy requirements. The more subspaces, the finer the division of working conditions and the higher the calculation accuracy, but the amount of calculation will increase. Those skilled in the art can choose reasonably according to actual engineering needs without excessive limitation.

[0022] The total load response at the corresponding target analysis section is calculated as follows: Using each type of working condition in the set of working conditions as a constraint, a time-domain simulation is performed on the finite element model. The logic for the time-domain simulation is as follows: In the simulation, a wind field based on average wind speed and wave direction is applied. The wind load response data at the target analysis section under wind action is obtained from the simulation results. The wind load response data consists of an axial force time series and a surrounding... Bending moment time series of the shaft and its surroundings The bending moment time series of the shaft together constitute the structure; In the simulation, a wave field based on significant wave height, spectral peak period, and wave direction is applied. Based on the simulation results, wave load response data at the target analysis section under different wave actions are obtained. The wave load response data also consists of axial force time series and circumferential... Bending moment time series of the shaft and its surroundings The bending moment time series of the shaft is constituted; The wind load response data and wave load response data under each working condition are added together to obtain the total load response at the target analysis section under each working condition.

[0023] In the above process, the originally complex wind-wave coupling problem is decomposed into two relatively independent physical processes that are solved separately, and then the combined effect is restored by superposition. The advantages of doing so are: it avoids the expensive fully coupled time-domain simulation of all working conditions, and only the wind field and wave field need to be simulated in the time domain to obtain the load response data for each working condition, which greatly reduces the amount of simulation calculation; at the same time, the time series of axial force, bending moment about the y-axis and bending moment about the z-axis are used as a unified response vector form, which facilitates subsequent linear superposition and correction processing, and provides a clear data foundation for subsequent accuracy compensation based on RMS correction factor, thus laying the foundation for obtaining a high-precision total load response while ensuring computational efficiency. Obtain axial force time series, around Bending moment time series of the shaft and its surroundings The bending moment time series of the shaft is because these three components fully describe the internal force state of the target analysis section under the combined action of complex wind and waves: the axial force reflects the tensile and compressive forces along the axis of the member, and the bending moment time series around the shaft is also included. Bending moment and circumference of shaft The bending moments of the shaft represent the bending action in two orthogonal directions, which together constitute all the information of the cross-sectional load response. Obtaining the time series of these three components can not only preserve the dynamic characteristics of the load changing over time, but also provide a complete input for subsequent stress transformation based on Euler-Bernoulli beam theory. , , The axis is determined based on the local coordinate system where the target analysis section is located: where The axis is always along the axial direction of the floating structure at the analysis section, that is, perpendicular to the plane of the analysis section. shaft and The axis lies within the plane of the analysis section and is typically set according to the direction of the structure's principal axis of inertia, such as in a circular or rectangular section. The axis can be set to the horizontal direction. The axes are set to the vertical direction, and the two axes are perpendicular to each other, forming a right-handed coordinate system.

[0024] S3: Calculate the probability of each working condition in the set of working conditions and determine the representative set of working conditions. Based on the representative set of working conditions, perform a fully coupled time-domain simulation of the entire floating structure of the floating wind turbine containing the target analysis section. Extract the fully coupled load response at the target analysis section from the simulation results as the verification working condition parameter. Calculate the RMS correction factor based on the verification working condition parameter and correct the total load response of the target analysis section under each working condition. The RMS correction factor is calculated as follows: The working conditions are sorted from high to low according to the probability weight of each working condition in the set of working conditions, and the working conditions with a cumulative probability of more than 80% are selected in turn to form a representative set of working conditions. Based on a representative set of operating conditions, a fully coupled time-domain simulation of the entire floating wind turbine system, including the target analysis section, is performed. The fully coupled load response at the target analysis section is extracted from the simulation results as verification operating condition parameters. Each verification operating condition parameter includes the fully coupled axial force time series and the load response around the target analysis section. Fully coupled bending moment time series of the shaft and around Fully coupled bending moment time series of the shaft; For each verification condition, the axial force time series under wind action and the axial force time series under wave action are first added together to obtain a linear superimposed axial force time series. The root mean square values ​​of the fully coupled axial force time series and the linear superimposed axial force time series are calculated respectively. The specific calculation method is as follows: the value at each moment in the time series is squared, all squared values ​​are integrated over the simulation time, the integration result is divided by the simulation time, and the square root of the quotient is taken to obtain the root mean square value of the fully coupled axial force. The root mean square value of the fully coupled axial force is divided by the root mean square value of the linearly superimposed axial force time series to obtain the axial force ratio for this verification condition. Using the same steps, the bending moments about the y-axis and z-axis are calculated separately to obtain the bending moment ratios about the y-axis and z-axis. The axial force ratios of all verification conditions are summed and divided by the total number of verification conditions to obtain the axial force RMS correction factor for the target analysis section. The bending moment ratios about the y-axis of all verification conditions are summed and divided by the total number of verification conditions to obtain the bending moment RMS correction factor about the y-axis. The bending moment ratios about the z-axis of all verification conditions are summed and divided by the total number of verification conditions to obtain the bending moment RMS correction factor about the z-axis.

[0025] The formula used in the above process is: Indicates the first One verification operating condition parameter; Indicates the first The time series of fully coupled axial forces at the target analysis section under a verification working condition; Indicates the first The analysis section of the verification working condition target is around Fully coupled bending moment time series of the shaft; Indicates the first The analysis section of the verification working condition target is around Fully coupled bending moment time series of the shaft; The formula used to calculate the RMS correction factor based on the verification operating condition parameters is logically as follows: in, , , These represent the first, second, and third RMS correction factors for the target analysis section, respectively. Indicates the total number of verification conditions; Indicates the duration of the fully coupled time-domain simulation; Indicates the first Axial force time series of the target analysis section under a verification working condition wind field; Indicates the first Axial force time series of the target analysis section under a wave field under a verification working condition; Indicates the first The target analysis section under the verification working condition wind field. Time series of bending moments of the shaft; Indicates the first The target analysis section under the wave field of the verification working condition is surrounded by... Time series of bending moments of the shaft; Indicates the first The target analysis section under the verification working condition wind field. Time series of bending moments of the shaft; Indicates the first The target analysis section under the wave field of the verification working condition is surrounded by... Time series of bending moments of the shaft.

[0026] In the above process, by selecting high-probability working conditions with a cumulative probability of over 80% as a representative set of working conditions, and calculating the RMS correction factor based on these representative working conditions through fully coupled time-domain simulation, the technical effect is to achieve accurate compensation for the wind-wave coupling effect with minimal cost of fully coupled simulation. The advantage of doing so is that only a small number of high-probability working conditions need to be subjected to expensive fully coupled simulation to capture the main characteristics of wind-wave coupling. By calculating the root mean square ratio of the fully coupled response and the linear superposition response and taking the average, the RMS correction factor applicable to the target analysis section is obtained. Then, this correction factor is applied to the linear superposition results of all working conditions, which avoids the huge computational burden of directly performing fully coupled simulation on all working conditions and can effectively correct the error caused by ignoring the coupling effect in the decoupled linear superposition of wind and waves. The theoretical basis for using the RMS correction factor is that the impact of wind-wave coupling on load response, in an energy statistical sense, is mainly manifested as a scaling of amplitude levels, rather than a fundamental change in phase and frequency components. Numerous marine engineering studies have shown that for load responses at the cross-sections of floating structures, such as axial force and bending moment, the main effect of wind-wave coupling is to enhance or weaken the overall energy level of the load, while the dominant frequency components of the response are primarily determined by the wave frequency; the coupling effect has a relatively small impact on the frequency distribution. Fully coupled time-domain simulations are typically implemented using specialized integrated simulation software or tools, such as the open-source software OpenFAST. These simulations perform a complete dynamic response numerical simulation of the entire floating wind turbine system, including the target analysis section, within the time domain. The equations of motion comprehensively consider structural mass, added mass, damping matrix, still-water restoring stiffness, and external loads under the combined action of wind and waves. By solving these equations, the motion response of the floating structure in a real marine environment and the load time histories at each analysis section can be obtained, including axial force, bending moment time series around the y-axis and z-axis. These fully coupled simulation results are considered the closest "benchmark values" to the actual situation and are used for subsequent comparison with the decoupled linear superposition results of wind and waves to calculate the RMS correction factor.

[0027] The corrected total load response is obtained as follows: First, the axial force time series of the wind field and wave field under this working condition are added together to obtain a linearly superimposed axial force time series. Then, the axial force RMS correction factor is multiplied by this linearly superimposed axial force time series to obtain the corrected axial force time series. For the bending moment about the y-axis, the y-axis bending moment time series of the wind field and wave field are added together to obtain a linearly superimposed y-axis bending moment time series. Then, the y-axis bending moment RMS correction factor is multiplied by this linearly superimposed y-axis bending moment time series to obtain the corrected y-axis bending moment time series. For the bending moment about the z-axis, the same steps are used, multiplying the z-axis bending moment RMS correction factor by the linearly superimposed z-axis bending moment time series to obtain the corrected z-axis bending moment time series. Finally, the corrected axial force time series, the corrected y-axis bending moment time series, and the corrected z-axis bending moment time series are combined to form the corrected total load response for this working condition.

[0028] The formula upon which the above process is based is: in, in, Indicates the first Corrected axial force at the target analysis section under each working condition; Indicates the first The corrected target analysis section under each working condition is then... Time series of bending moments of the shaft; Indicates the first After correction of the target analysis section under each working condition, it is surrounded by... Time series of bending moments of the shaft; Indicates the first The total load response of the target analysis section after the working condition correction.

[0029] In the above process, the corrected total load response is obtained by multiplying the RMS correction factors α, β, and γ calculated based on representative working conditions by the time series of axial force, bending moment around the y-axis, and bending moment around the z-axis after the linear superposition of wind load and wave load under each working condition. The technical effect is to achieve accurate compensation for wind-wave coupling effect with minimal computational cost. The correction factors are statistically obtained based on the full coupling simulation results of a small number of high-probability working conditions, which can accurately reflect the influence law of wind-wave coupling on load response. After applying them to the linear superposition results of all working conditions, the efficiency of wind-wave decoupling method is retained, and the error caused by ignoring coupling effect is effectively corrected, so that the corrected total load response is highly close to the real full coupling response in terms of energy statistics.

[0030] S4: Based on the geometric characteristics of the target analysis section, the Euler-Bernoulli beam method is used to convert the corrected total load response into local stress. Fourier transform analysis is performed on the local stress to calculate the short-term fatigue damage coefficient in the frequency domain. Based on the local stress and the rainflow count in the time domain, and combined with the SN curve, the short-term fatigue damage coefficient in the time domain is calculated. The results of the short-term fatigue damage coefficient in the frequency domain and the short-term fatigue damage coefficient in the time domain are compared to select the optimal stress amplitude probability density function for the target analysis section under each working condition. The calculation of the short-term fatigue damage coefficient in the frequency domain is as follows: Divide each value in the corrected axial force time series of the target analysis section under each working condition by the cross-sectional area of ​​that section to obtain the stress time series generated by the axial force; take the absolute value of each value in the corrected bending moment time series of the target analysis section about the y-axis under each working condition and divide it by the section modulus of that section about the y-axis to obtain the stress time series generated by bending about the y-axis; take the absolute value of each value in the corrected bending moment time series of the target analysis section about the z-axis under each working condition and divide it by the section modulus of that section about the z-axis to obtain the stress time series generated by bending about the z-axis; add the values ​​of these three stress time series at corresponding times to obtain the local stress time series of the target analysis section under each working condition. The stress power spectrum is obtained by performing a Fourier transform on the local stress time series. The specific calculation method is as follows: For each frequency value, multiply each value in the local stress time series by an exponential function with the natural constant as the base and negative imaginary units multiplied by the frequency and the corresponding time. Summate all the products within the range of zero to the simulation duration to obtain the Fourier transform result at that frequency. Take the square of the modulus of the Fourier transform result and divide it by the simulation duration. When the simulation duration approaches infinity, this value is the stress power spectrum value at that frequency. For the zeroth spectral moment, the stress power spectrum is integrated over a frequency range from zero to infinity, i.e., the stress power spectrum values ​​corresponding to all frequencies are multiplied by the frequency interval and then summed. For the first spectral moment, each frequency value is multiplied by the stress power spectrum value at that frequency, and then integrated over the frequency range. For the second spectral moment, the square of the frequency is multiplied by the stress power spectrum value, and then integrated. For the fourth spectral moment, the fourth power of the frequency is multiplied by the stress power spectrum value, and then integrated, thus obtaining the zeroth, first, second, and fourth spectral moments, respectively. Based on the calculated spectral moments, four different stress amplitude probability density function models were compared, including the Rayleigh model, Dirlik model, Wirsching model, and Zhao-Baker model. For each probability density function model, short-term fatigue damage in the frequency domain was calculated. The specific calculation method was as follows: the stress range was used as the integration variable. First, the slope of the SN curve of the stress range was calculated to the power of the slope. Then, it was multiplied by the probability density of the stress range under the model. The product was integrated over the stress range from zero to infinity to obtain the integral result. The integral result was multiplied by the simulation duration and then divided by the material constant of the SN curve to obtain the short-term fatigue damage coefficient in the frequency domain under the model.

[0031] The formula upon which the above process is based is: in, Indicates the first Local stress in the target analysis section under various working conditions; This represents the cross-sectional area of ​​the target analysis section; Indicates the target analysis section around Section modulus of the shaft; Indicates the target analysis section around Section modulus of the shaft; In the above process, the dependent variable Indicates the first The local stress time history of the target analysis section under various working conditions provides core input data for subsequent frequency domain fatigue damage calculation by quantifying the actual stress level of the dangerous part under the combined action of wind and waves; independent variables , and Representing the corrected axial force and the circumferential force, respectively. , The time history of bending moment of the shaft, these load components and cross-sectional geometry , , The magnitude of local stress is jointly determined by the axial force and its physical relationship is based on the Euler-Bernoulli beam theory; the stress generated by the axial force in the formula is related to the stress generated by the axial force and the stress generated by the axial force. The bending stress generated by the bending moment is directly proportional to the bending stress generated by the bending moment. and They are directly proportional, therefore the local stress It is positively correlated with all three load variables; an increase in any load component will lead to a corresponding increase in local stress.

[0032] Performing a Fourier transform on the local stresses yields the stress power spectrum of each target analysis section: in, Indicates the first Stress power spectrum of the target analysis section under various working conditions; Indicates the duration of the time-domain simulation; The frequency represents the stress power spectrum; Calculate the spectral moments of each order of the stress power spectrum: in, , indicating the order of the spectral moment; Indicates the first The first target analysis section under each working condition Spectral moments; Based on spectral moments, different known stress amplitude probability density function models are compared, and short-term fatigue damage in the frequency domain is calculated for each model. These models include the Rayleigh, Dirlik, Wirsching, and Zhao-Bzker probability density function models. in, Indicates the first The target analysis section under each working condition is at the first... Short-term fatigue damage coefficient in the frequency domain under a stress amplitude probability density function model; Indicates the first Under the first working condition The known first analysis section Stress amplitude probability density function model; Indicates the material constants of the SN curve; Indicates the stress range; Indicates the index of the number of models for the stress amplitude probability density function; This represents the slope of the SN curve.

[0033] In the above process, the dependent variable Indicates the first The target analysis section under each working condition is at the first The short-term fatigue damage coefficient in the frequency domain under a stress amplitude probability density function model quantifies the degree of fatigue damage accumulation per unit time under specific working conditions using a frequency domain method, providing basic data for the subsequent weighted integral of long-term fatigue damage; independent variables include simulation duration. SN curve material constant Stress range SN curve slope and stress amplitude probability density function ,in and They are directly proportional. and They are inversely proportional. As a multiplier in the integrand, it reflects the nonlinear amplification effect of high stress range on damage, and This determines the weighting of different stress ranges in the total damage; through the integral operation of this formula, the statistical characteristics of the random stress process can be organically combined with the fatigue performance parameters of the material, realizing an efficient conversion from stress power spectrum to fatigue damage. By converting the corrected total load response into local stress, and then performing a Fourier transform to obtain the stress power spectrum, and calculating the short-term fatigue damage coefficient in the frequency domain based on the spectral moment, the technical advantage is that it transforms the complex stress cycle counting problem in the time domain into an efficient integral calculation problem in the frequency domain. This avoids the time-consuming rainflow counting of long-term stress histories for each working condition, and instead uses the power spectral density function and spectral moment parameters, combined with the stress amplitude probability density function model, to quickly calculate the frequency domain fatigue damage. At the same time, the bandwidth coefficient based on the spectral moment can accurately describe the broadband characteristics of the stress response, providing a basis for subsequent comparison of different stress amplitude probability density function models. The slope of the SN curve specifically refers to the inclination of the curve describing the relationship between stress range and fatigue life in a logarithmic coordinate system. It reflects the sensitivity of the material to changes in stress range on the number of cycles it can withstand. A larger slope means that a small increase in stress range will lead to a significant decrease in the material's allowable number of cycles, indicating that the material is more sensitive to high stress ranges; a smaller slope means that changes in stress range have a relatively gentle effect on fatigue life.

[0034] The Rayleigh probability density function model is as follows: Represented as zeroth order spectral moment; Represents the Rayleigh probability density function model; The Dirlik probability density function model is: in, parameter These represent the first, second, and third indices of the model, respectively. These represent the first and second shape parameters of the model, respectively. Indicates the standardized stress range; This represents the Dirlik probability density function model; The Wirsching probability density function model is: in, , This is an empirical coefficient relating to the slope of the SN curve; This is the bandwidth factor; Represents the Wirsching probability density function model; The Zhao-Bzker probability density function model is: in, in, This represents the modified Rayleigh probability density function model; The correction factor for the spectral moment; These represent the weighting coefficients of the Rayleigh probability density function model; This represents the Zhao-Bzker probability density function model; The comparison of these probability density function models is due to the significant differences in the power spectrum shape and bandwidth characteristics of stress response under different operating conditions: the Rayleigh probability density function model is only applicable to narrowband processes, and for broadband processes, it overestimates the probability of large stress ranges, resulting in overly conservative results; the Dirlik probability density function model is obtained by fitting a large amount of simulation data and has the best adaptability to broadband processes, but its parameter calculation is complex; the Wirsching probability density function model extends the applicability of the narrowband method through correction factors; and the Zhao-Baker probability density function model further improves flexibility through weighted combination. No single model is universally applicable to all operating conditions. Therefore, by comparing the relative errors between the frequency domain damage calculated by each probability density function model and the time domain rainflow counting benchmark results, the probability density function model that best matches the response characteristics of each operating condition can be adaptively selected, thereby maximizing the accuracy of fatigue damage calculation while maintaining the efficiency of the frequency domain method.

[0035] The optimal stress amplitude probability density function is selected as follows: Based on local stress and time-domain rainflow counting, and combined with SN curves, the short-term fatigue damage coefficient of the target analysis section under each working condition is calculated as follows: the peak and valley values ​​in the local stress time series are extracted sequentially and paired according to the rules of rainflow counting to identify complete stress cycles; for each identified stress cycle, the difference between its maximum and minimum values ​​is calculated to obtain the stress range of the cycle; all stress cycles are grouped according to the size of the stress range, and the number of cycles contained in each stress range is counted to obtain the value of each stress range and the number of times the stress range appears; The allowable number of cycles for each stress range is calculated based on the SN curve. The specific calculation method is as follows: divide the reference plate thickness by the actual plate thickness and calculate the ratio between the two; multiply the ratio by the thickness effect exponent to obtain the thickness correction factor; multiply each stress range by the thickness correction factor to obtain the corrected stress range; take the negative power of the slope of the SN curve for the corrected stress range and multiply it by the intercept of the SN curve to obtain the allowable number of cycles for that stress range. For each stress range, divide the number of times the stress range appears by the corresponding allowable number of cycles to obtain the fatigue damage caused by the stress range. Add up the fatigue damage of all groups to obtain the time-domain short-term fatigue damage coefficient of the target analysis section under each working condition. The formula upon which the above process is based is: in, in, Indicates the first The total number of stress cycles at the target analysis section under each working condition; Indicates stress range Number of times it appears; Indicates the first Group stress range; Indicates stress range The corresponding number of allowed loops; Indicates plate thickness; Indicates the reference plate thickness; Indicates the thickness effect index; Indicates the slope of the SN curve; Indicates the intercept of the SN curve; Indicates the first Short-term fatigue damage coefficient of the target analysis section under various working conditions in the time domain; In the above process, the dependent variable Indicates the first The time-domain short-term fatigue damage coefficient of the target analysis section under each working condition is directly reflected by rainflow counting and the linear cumulative damage criterion, indicating the degree of fatigue damage accumulation per unit time under this condition. This provides a benchmark true value for subsequent comparison with frequency domain damage and selection of the optimal stress amplitude probability density function. The independent variables include the stress range for each group. Number of times The allowable number of cycles for each stress range , plate thickness Reference plate thickness Thickness effect index SN curve slope and the intercept of the SN curve ,in and They are directly proportional. and They are inversely proportional. By influencing Indirectly determines the size of the damage, and , , , , Determined by both material and geometric parameters The specific numerical values; through the discrete summation operation of this formula, the complex stress time history can be decomposed into several typical stress cycles and their damage contribution can be quantified, providing a reliable time domain benchmark for evaluating the accuracy of different probability density function models in the frequency domain method; All complete stress cycles were extracted from the local stress time history by rainflow counting, and statistics were compiled for each stress range. Number of times Then, by combining the SN curve, the allowable number of cycles for each stress range can be calculated. Finally, the damage ratios of all cycles are summed according to the Palmgern-Miner linear cumulative damage criterion to obtain the short-term fatigue damage coefficient in the time domain. The technical effect is to obtain the benchmark true value of fatigue damage in the most direct way in the time domain. The advantage of doing so is that rainflow counting, as an internationally recognized standard method for extracting stress cycles, can truly reflect the cyclic characteristics in the stress time history. The damage results obtained after combining with the SN curve have the highest reliability. Therefore, it is used as a benchmark for subsequent frequency domain damage calculation to evaluate the accuracy of different frequency domain methods and select the best stress amplitude probability density function model accordingly. The short-term frequency domain fatigue damage results are compared with the short-term time domain fatigue damage coefficient results to calculate the relative error. The stress amplitude probability density function model that minimizes the relative error is selected as the optimal stress amplitude probability density function.

[0036] In the above process, by comparing the short-term frequency domain fatigue damage with the short-term time domain fatigue damage coefficient, the relative error of each stress amplitude probability density function model is calculated, and the model with the smallest error is selected as the optimal stress amplitude probability density function. The technical effect is to achieve adaptive matching of the frequency domain fatigue analysis method to the working condition characteristics.

[0037] S5: Based on the optimal stress amplitude probability density function, recalculate the frequency domain short-term fatigue damage coefficient for all working conditions to obtain the optimal short-term fatigue damage coefficient: Integrate the optimal short-term fatigue damage under all working conditions, and combine it with the probability under each working condition to calculate the long-term fatigue damage coefficient and evaluate the fatigue life.

[0038] The fatigue life assessment is conducted as follows: Based on the optimal stress amplitude probability density function, the frequency domain short-term fatigue damage coefficient is recalculated for all working conditions to obtain the optimal short-term fatigue damage coefficient. The stress range is used as the integration variable. First, the slope parameter of the SN curve of the stress range is calculated and then multiplied by the probability density of the stress range given by the optimal stress amplitude probability density function under this working condition. The product is integrated over the stress range from zero to infinity to obtain the integration result. The integration result is multiplied by the simulation duration and then divided by the SN curve constant to obtain the optimal short-term fatigue damage coefficient of the target analysis section under this working condition. The optimal short-term fatigue damage coefficients for all working conditions are summed, and the long-term fatigue damage coefficient is calculated by combining the probability of each working condition and the ratio of the design life to the simulation duration. Specifically, the optimal short-term fatigue damage coefficient for each working condition is multiplied by the probability of that working condition, and then multiplied by the quotient of the design life divided by the simulation duration to obtain the contribution of that working condition to long-term fatigue damage. The contributions of all working conditions are then summed to obtain the long-term fatigue damage coefficient of the target analysis section. The fatigue life of the target analysis section is assessed based on the long-term fatigue damage coefficient. The specific calculation method is as follows: divide the value by the long-term fatigue damage coefficient, and the quotient is the fatigue life of the target analysis section.

[0039] The formula upon which the above process is based is: in, Indicates the first The optimal short-term fatigue damage coefficient of the target analysis section under each working condition; Indicates the first The optimal stress amplitude probability density function for the target analysis section under each working condition; Indicates stress range of Power; Indicates the slope of the SN curve; In the above process, the dependent variable Indicates the first The optimal short-term fatigue damage coefficient, calculated based on the optimal stress amplitude probability density function, is obtained from the target analysis section under each working condition. This parameter quantifies the degree of fatigue damage accumulation per unit time of the structure under this working condition using the frequency domain integral method, providing accurate basic data for the subsequent weighted summation of long-term fatigue damage. Independent variables include simulation duration. SN curve material constant Stress range SN curve slope and the optimal stress amplitude probability density function ,in and They are directly proportional. and They are inversely proportional. The multiplier in the integrand reflects the nonlinear amplification effect of high stress on damage, while This determines the weighting of different stress ranges in the total damage; The optimal stress amplitude probability density function, verified through error comparison under each working condition, was adopted. The optimal short-term fatigue damage coefficient was obtained by recalculating the frequency domain integral for all operating conditions. Its technical advantages are the precise preservation and efficient calculation of frequency domain fatigue analysis. On the one hand, it avoids the huge amount of calculation required by traditional time domain rainflow counting to extract each working condition one by one. On the other hand, it ensures a high degree of consistency between the frequency domain calculation results and the time domain reference value by using the optimal stress amplitude probability density function model selected through prior comparison.

[0040] Integrate the optimal short-term fatigue damage coefficient for all working conditions, and combine this with the probability for each working condition to calculate the long-term fatigue damage coefficient: in, This represents the long-term fatigue damage coefficient of the target analysis section; Indicates the design life of the entire wind turbine; Indicates the first The probability of each working condition; In the above process, the optimal short-term fatigue damage coefficient for each working condition is determined. Multiply by the probability of this working condition occurring. The ratio of design life to simulation duration The long-term fatigue damage coefficient of the target analysis section is obtained by summing up all working conditions. The technical effect is to achieve accurate extrapolation from short-term damage to damage throughout the entire life cycle. The advantage of doing so is that, by using probability weighting, it truly reflects the actual contribution ratio of different sea conditions during the entire service life of the wind turbine, avoiding the bias caused by equal weighted averaging. At the same time, the extrapolation of the time scale reasonably extends the short-term simulation results to the long-term design life, so that the final assessment of long-term damage can comprehensively cover all environmental conditions that the wind turbine may encounter during its life cycle, providing a reliable mathematical basis for fatigue life assessment.

[0041] Based on the long-term fatigue damage coefficient, fatigue life is assessed: in, This indicates the fatigue life of the target analysis section.

[0042] In the above process, the fatigue life of the target analysis section is obtained by calculating the reciprocal of the long-term fatigue damage coefficient. The technical effect is to transform the abstract amount of accumulated damage into an intuitive assessment of service life. The advantage of doing so is that, based on the Palmgren-Miner linear cumulative damage criterion, the structure will fail due to fatigue when the accumulated damage reaches 1. Therefore, the reciprocal of the long-term damage coefficient directly reflects the total time that the structure can operate safely under the current environmental load.

[0043] In the above embodiments, 20 sets of data on the long-term fatigue damage coefficient and the corresponding fatigue life of the target analysis section are given to reflect the change of the fatigue life of the target analysis section with the long-term fatigue damage coefficient, as shown in Table 1: Table 1: Relationship between long-term fatigue damage coefficient and fatigue life of the corresponding target analysis section In Table 1 above, the fatigue life of the target analysis section decreases with the increase of the long-term fatigue damage coefficient, which is consistent with the changing trend of the fatigue life of the structure under the current environment.

[0044] Please see Figure 3 The present invention also provides a fatigue damage analysis system for the floating body structure of a floating wind turbine, the analysis system being used to perform the above-described analysis method, including: Data acquisition module: used to construct the finite element model of the floating structure to be analyzed, and to label the cross-sections between adjacent structural segments within the floating structure as analysis cross-sections, and to determine the target analysis cross-section from the analysis cross-sections; Total load calculation module: used to obtain several types of wind and wave parameters in the historical area of ​​the floating structure deployment area. The wind and wave parameters are cut by recursive iteration to construct a set of working conditions. Each type of working condition in the set of working conditions is used as a constraint condition to perform time-domain simulation on the finite element model to obtain the total load response at the target analysis section. Correction module: Calculates the probability of each working condition in the set of working conditions and determines the representative set of working conditions. Based on the representative set of working conditions, it performs a fully coupled time-domain simulation of the entire floating structure of the floating wind turbine containing the target analysis section. It extracts the fully coupled load response at the target analysis section from the simulation results as the verification working condition parameter. Based on the verification working condition parameter, it calculates the RMS correction factor and corrects the total load response of the target analysis section under each working condition. The Fourier transform module is used to convert the corrected total load response into local stress based on the geometric characteristics of the target analysis section using the Euler-Bernoulli beam method. Fourier transform analysis is performed on the local stress to calculate the short-term fatigue damage coefficient in the frequency domain. Based on the local stress and the rainflow count in the time domain, and combined with the SN curve, the short-term fatigue damage coefficient in the time domain is calculated. The results of the short-term fatigue damage coefficient in the frequency domain are compared with those in the time domain to select the optimal stress amplitude probability density function for the target analysis section under each working condition. Life assessment module: It is used to recalculate the frequency domain short-term fatigue damage coefficient for all working conditions based on the optimal stress amplitude probability density function to obtain the optimal short-term fatigue damage coefficient. It integrates the optimal short-term fatigue damage under all working conditions and combines it with the probability under each working condition to calculate the long-term fatigue damage coefficient and assess fatigue life.

[0045] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0046] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0047] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0048] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for fatigue damage analysis of the floating body structure of a floating wind turbine, characterized in that, include: S1: Construct a finite element model of the floating structure to be analyzed, and label the cross-sections between adjacent structural segments within the floating structure as analysis cross-sections, and determine the target analysis cross-section from the analysis cross-sections; S2: Obtain several types of wind and wave parameters from the historical data of the floating structure deployment area, and use recursive iterative cutting to cut the wind and wave parameters to construct a set of working conditions. Using each type of working condition in the set of working conditions as a constraint, perform time-domain simulation on the finite element model to obtain the total load response at the target analysis section. S3: Calculate the probability of each working condition in the set of working conditions and determine the representative set of working conditions. Based on the representative set of working conditions, perform a fully coupled time-domain simulation of the entire floating structure of the floating wind turbine containing the target analysis section. Extract the fully coupled load response at the target analysis section from the simulation results as the verification working condition parameter. Calculate the RMS correction factor based on the verification working condition parameter and correct the total load response of the target analysis section under each working condition. S4: Based on the geometric characteristics of the target analysis section, the Euler-Bernoulli beam method is used to convert the corrected total load response into local stress. Fourier transform analysis is performed on the local stress to calculate the short-term fatigue damage coefficient in the frequency domain. Based on the local stress and the rainflow count in the time domain, and combined with the SN curve, the short-term fatigue damage coefficient in the time domain is calculated. The results of the short-term fatigue damage coefficient in the frequency domain and the short-term fatigue damage coefficient in the time domain are compared to select the optimal stress amplitude probability density function for the target analysis section under each working condition. S5: Based on the optimal stress amplitude probability density function, recalculate the frequency domain short-term fatigue damage coefficient for all working conditions to obtain the optimal short-term fatigue damage coefficient: Integrate the optimal short-term fatigue damage under all working conditions, and combine it with the probability under each working condition to calculate the long-term fatigue damage coefficient and evaluate the fatigue life. The set of working conditions is constructed by recursive iterative cutting, specifically as follows: Several types of wind and wave parameters from the historical data of the floating structure deployment area are obtained, including average wind speed, significant wave height, spectral peak period, and wind and wave direction. These parameters are then labeled as a space of data points representing a four-dimensional parameter combination. After sorting all observations for each dimension, the midpoint between each adjacent observation point is sequentially considered as a candidate cutting point. The number of data points contained in the left and right subspaces after the cutting is calculated, and the cutting point that makes the number of points in the left and right subspaces closest to equal is selected as the optimal cutting point for that dimension. The optimal cutting results in the four dimensions are compared, and the dimension with the smallest difference in the number of data points in the left and right subspaces and its corresponding cutting point are selected for actual cutting, thus dividing the current space into two. Repeat the above process for each newly generated subspace, recursively cutting it until the number of subspaces obtained reaches the preset number. These final subspaces are the equally probable working condition intervals. The average value of each type of dimension parameter within each working condition interval is marked as the final parameter of the working condition interval, forming a discrete set of working conditions.

2. The method for fatigue damage analysis of the floating body structure of a floating wind turbine according to claim 1, characterized in that, The total load response at the corresponding target analysis section is calculated as follows: Using each type of working condition in the set of working conditions as a constraint, a time-domain simulation is performed on the finite element model. The logic for the time-domain simulation is as follows: In the simulation, a wind field based on average wind speed and wave direction is applied. The wind load response data at the target analysis section under wind action is obtained from the simulation results. The wind load response data consists of an axial force time series and a surrounding... Bending moment time series of the shaft and its surroundings The bending moment time series of the shaft together constitute the structure; In the simulation, a wave field based on significant wave height, spectral peak period, and wave direction is applied. Based on the simulation results, wave load response data at the target analysis section under different wave actions are obtained. The wave load response data also consists of axial force time series and... Bending moment time series of the shaft and its surroundings The bending moment time series of the shaft is constituted; The wind load response data and wave load response data under each working condition are added together to obtain the total load response at the target analysis section under each working condition.

3. The method for fatigue damage analysis of the floating body structure of a floating wind turbine according to claim 1, characterized in that, The RMS correction factor is calculated as follows: The working conditions are sorted from high to low according to the probability weight of each working condition in the set of working conditions, and the working conditions with a cumulative probability of more than 80% are selected in turn to form a representative set of working conditions. Based on a representative set of operating conditions, a fully coupled time-domain simulation of the entire floating wind turbine system, including the target analysis section, is performed. The fully coupled load response at the target analysis section is extracted from the simulation results as verification operating condition parameters. Each verification operating condition parameter includes the fully coupled axial force time series and the load response around the target analysis section. Fully coupled bending moment time series of the shaft and around Fully coupled bending moment time series of the shaft; For each verification condition, the axial force time series under wind action and the axial force time series under wave action are first added together to obtain a linear superimposed axial force time series. The root mean square values ​​of the fully coupled axial force time series and the linear superimposed axial force time series are calculated respectively. The specific calculation method is as follows: the value at each moment in the time series is squared, all squared values ​​are integrated over the simulation time, the integration result is divided by the simulation time, and the square root of the quotient is taken to obtain the root mean square value of the fully coupled axial force. The root mean square value of the fully coupled axial force is divided by the root mean square value of the linearly superimposed axial force time series to obtain the axial force ratio for this verification condition. Using the same steps, the bending moments about the y-axis and z-axis are calculated separately to obtain the bending moment ratios about the y-axis and z-axis. The axial force ratios of all verification conditions are summed and divided by the total number of verification conditions to obtain the axial force RMS correction factor for the target analysis section. The bending moment ratios about the y-axis of all verification conditions are summed and divided by the total number of verification conditions to obtain the bending moment RMS correction factor about the y-axis. The bending moment ratios about the z-axis of all verification conditions are summed and divided by the total number of verification conditions to obtain the bending moment RMS correction factor about the z-axis.

4. The method for fatigue damage analysis of the floating body structure of a floating wind turbine according to claim 3, characterized in that, The corrected total load response is obtained as follows: First, the axial force time series of the wind field and wave field under this working condition are added together to obtain a linearly superimposed axial force time series. Then, the axial force RMS correction factor is multiplied by this linearly superimposed axial force time series to obtain the corrected axial force time series. For the bending moment about the y-axis, the y-axis bending moment time series of the wind field and wave field are added together to obtain a linearly superimposed y-axis bending moment time series. Then, the y-axis bending moment RMS correction factor is multiplied by this linearly superimposed y-axis bending moment time series to obtain the corrected y-axis bending moment time series. For the bending moment about the z-axis, the same steps are used, multiplying the z-axis bending moment RMS correction factor by the linearly superimposed z-axis bending moment time series to obtain the corrected z-axis bending moment time series. Finally, the corrected axial force time series, the corrected y-axis bending moment time series, and the corrected z-axis bending moment time series are combined to form the corrected total load response for this working condition.

5. The method for fatigue damage analysis of the floating body structure of a floating wind turbine according to claim 4, characterized in that, The calculation of the short-term fatigue damage coefficient in the frequency domain is as follows: Divide each value in the corrected axial force time series of the target analysis section under each working condition by the cross-sectional area of ​​that section to obtain the stress time series generated by the axial force; take the absolute value of each value in the corrected bending moment time series of the target analysis section about the y-axis under each working condition and divide it by the section modulus of that section about the y-axis to obtain the stress time series generated by bending about the y-axis; take the absolute value of each value in the corrected bending moment time series of the target analysis section about the z-axis under each working condition and divide it by the section modulus of that section about the z-axis to obtain the stress time series generated by bending about the z-axis; add the values ​​of these three stress time series at corresponding times to obtain the local stress time series of the target analysis section under each working condition. The stress power spectrum is obtained by performing a Fourier transform on the local stress time series. The specific calculation method is as follows: For each frequency value, multiply each value in the local stress time series by an exponential function with the natural constant as the base and negative imaginary units multiplied by the frequency and the corresponding time. Summate all the products within the range of zero to the simulation duration to obtain the Fourier transform result at that frequency. Take the square of the modulus of the Fourier transform result and divide it by the simulation duration. When the simulation duration approaches infinity, this value is the stress power spectrum value at that frequency. For the zeroth spectral moment, the stress power spectrum is integrated over a frequency range from zero to infinity, i.e., the stress power spectrum values ​​corresponding to all frequencies are multiplied by the frequency interval and then summed. For the first spectral moment, each frequency value is multiplied by the stress power spectrum value at that frequency, and then integrated over the frequency range. For the second spectral moment, the square of the frequency is multiplied by the stress power spectrum value, and then integrated. For the fourth spectral moment, the fourth power of the frequency is multiplied by the stress power spectrum value, and then integrated, thus obtaining the zeroth, first, second, and fourth spectral moments, respectively. Based on the calculated spectral moments, four different stress amplitude probability density function models were compared, including the Rayleigh model, Dirlik model, Wirsching model, and Zhao-Baker model. For each probability density function model, short-term fatigue damage in the frequency domain was calculated. The specific calculation method was as follows: the stress range was used as the integration variable. First, the slope of the SN curve of the stress range was calculated to the power of the slope. Then, it was multiplied by the probability density of the stress range under the model. The product was integrated over the stress range from zero to infinity to obtain the integral result. The integral result was multiplied by the simulation duration and then divided by the material constant of the SN curve to obtain the short-term fatigue damage coefficient in the frequency domain under the model.

6. The method for fatigue damage analysis of the floating body structure of a floating wind turbine according to claim 5, characterized in that, The optimal stress amplitude probability density function is selected as follows: Based on local stress and time-domain rainflow counting, and combined with SN curves, the short-term fatigue damage coefficient of the target analysis section under each working condition is calculated as follows: the peak and valley values ​​in the local stress time series are extracted sequentially and paired according to the rules of rainflow counting to identify complete stress cycles; for each identified stress cycle, the difference between its maximum and minimum values ​​is calculated to obtain the stress range of the cycle; all stress cycles are grouped according to the size of the stress range, and the number of cycles contained in each stress range is counted to obtain the value of each stress range and the number of times the stress range appears; The allowable number of cycles for each stress range is calculated based on the SN curve. The specific calculation method is as follows: divide the reference plate thickness by the actual plate thickness and calculate the ratio between the two. Multiply the ratio by the thickness effect exponent to obtain the thickness correction factor; multiply each stress range by the thickness correction factor to obtain the corrected stress range; take the negative of the slope of the SN curve to the power of the corrected stress range, and then multiply it by the intercept of the SN curve to obtain the allowable number of cycles corresponding to that stress range. For each stress range, divide the number of times the stress range appears by the corresponding allowable number of cycles to obtain the fatigue damage caused by the stress range. Add up the fatigue damage of all groups to obtain the time-domain short-term fatigue damage coefficient of the target analysis section under each working condition. The short-term frequency domain fatigue damage results are compared with the short-term time domain fatigue damage coefficient results to calculate the relative error. The stress amplitude probability density function model that minimizes the relative error is selected as the optimal stress amplitude probability density function.

7. The method for fatigue damage analysis of the floating body structure of a floating wind turbine according to claim 6, characterized in that, The fatigue life assessment is conducted as follows: Based on the optimal stress amplitude probability density function, the frequency domain short-term fatigue damage coefficient is recalculated for all working conditions to obtain the optimal short-term fatigue damage coefficient. The stress range is used as the integration variable. First, the slope parameter of the SN curve of the stress range is calculated and then multiplied by the probability density of the stress range given by the optimal stress amplitude probability density function under this working condition. The product is integrated over the stress range from zero to infinity to obtain the integration result. The integration result is multiplied by the simulation duration and then divided by the SN curve constant to obtain the optimal short-term fatigue damage coefficient of the target analysis section under this working condition. The optimal short-term fatigue damage coefficients for all working conditions are summed, and the long-term fatigue damage coefficient is calculated by combining the probability of each working condition and the ratio of the design life to the simulation duration. Specifically, the optimal short-term fatigue damage coefficient for each working condition is multiplied by the probability of that working condition, and then multiplied by the quotient of the design life divided by the simulation duration to obtain the contribution of that working condition to long-term fatigue damage. The contributions of all working conditions are then summed to obtain the long-term fatigue damage coefficient of the target analysis section. The fatigue life of the target analysis section is assessed based on the long-term fatigue damage coefficient. The specific calculation method is as follows: divide the value by the long-term fatigue damage coefficient, and the quotient is the fatigue life of the target analysis section.

8. A fatigue damage analysis system for the floating body structure of a floating wind turbine, wherein the analysis system is used to implement the analysis method described in claims 1-8, characterized in that: Data acquisition module: used to construct the finite element model of the floating structure to be analyzed, and to label the cross-sections between adjacent structural segments within the floating structure as analysis cross-sections, and to determine the target analysis cross-section from the analysis cross-sections; Total load calculation module: used to obtain several types of wind and wave parameters in the historical area of ​​the floating structure deployment area. The wind and wave parameters are cut by recursive iteration to construct a set of working conditions. Each type of working condition in the set of working conditions is used as a constraint condition to perform time-domain simulation on the finite element model to obtain the total load response at the target analysis section. Correction module: Calculates the probability of each working condition in the set of working conditions and determines the representative set of working conditions. Based on the representative set of working conditions, it performs a fully coupled time-domain simulation of the entire floating structure of the floating wind turbine containing the target analysis section. It extracts the fully coupled load response at the target analysis section from the simulation results as the verification working condition parameter. Based on the verification working condition parameter, it calculates the RMS correction factor and corrects the total load response of the target analysis section under each working condition. The Fourier transform module is used to convert the corrected total load response into local stress based on the geometric characteristics of the target analysis section using the Euler-Bernoulli beam method. Fourier transform analysis is performed on the local stress to calculate the short-term fatigue damage coefficient in the frequency domain. Based on the local stress and the rainflow count in the time domain, and combined with the SN curve, the short-term fatigue damage coefficient in the time domain is calculated. The results of the short-term fatigue damage coefficient in the frequency domain are compared with those in the time domain to select the optimal stress amplitude probability density function for the target analysis section under each working condition. Life assessment module: It is used to recalculate the frequency domain short-term fatigue damage coefficient for all working conditions based on the optimal stress amplitude probability density function to obtain the optimal short-term fatigue damage coefficient. It integrates the optimal short-term fatigue damage under all working conditions and combines it with the probability under each working condition to calculate the long-term fatigue damage coefficient and assess fatigue life.

Images