Structural optimization method and system of offshore converter station based on analogue simulation

Abstract

The invention provides a structural optimization method and system for an offshore converter station based on analogue simulation, and the method comprises the steps: carrying out the corrosion analogue simulation of a target converter station through a virtual simulation platform according to the service environment data of the target converter station, and obtaining an analogue simulation data set in a service environment; according to the analogue simulation data set, determining dominant influence factors influencing the corrosion rate through a preset multi-field coupling analysis model; according to the dominant influence factor and the analogue simulation data set, utilizing a time sequence analysis method to obtain distribution rule data of an erosion pit depth threshold value under tidal period change; generating a failure probability distribution diagram of the target converter station according to the distribution rule data; and according to the failure probability distribution diagram, generating a dynamic protection parameter set based on a stress corrosion coefficient and a fatigue corrosion factor, and performing structure optimization on the target converter station based on the dynamic protection parameter set, thereby improving the operation safety and reliability of the offshore converter station.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of simulation and high-strength steel corrosion resistance, and particularly relates to a structure optimization method and system for a marine converter station based on simulation. BACKGROUND

[0002] The long-term service reliability of offshore engineering equipment is crucial for energy transmission and marine resource development, especially in key facilities such as marine converter stations. High-strength steel, as the main structural material, directly affects the safety and service life of the equipment. The marine environment has characteristics of high temperature, high humidity, salt spray, and complex stress loading. These factors together accelerate the corrosion of high-strength steel and threaten the stable operation of the equipment. Existing corrosion prevention measures rely mainly on single coating protection or material modification, which cannot effectively address the complex corrosion mechanism in a multi-field coupled environment, especially under dynamic service conditions. Coating peeling and material performance degradation problems have not been effectively solved. In addition, existing research lacks quantitative analysis of the micro-mechanism of the corrosion process and the degradation of mechanical properties, resulting in a lack of scientific basis for corrosion prevention scheme design.

[0003] In the marine environment, the corrosion behavior of high-strength steel is influenced by the synergistic effect of temperature, stress, and corrosive medium. These factors interact with each other and jointly determine the corrosion rate and morphology evolution. The interaction of temperature and medium accelerates the formation of pits, while the presence of stress further promotes the expansion of pits, leading to an increased risk of local failure of the material. The quantitative relationship between the dynamic evolution of pit size and the corrosion period is not clear, limiting the prediction ability of the corrosion process. At the same time, the correlation between the mechanical property degradation of high-strength steel after pre-corrosion and the microstructure change is insufficient, making the optimization of corrosion-resistant coatings and alloy compositions for marine environments lack targeted guidance. Therefore, the key problem is to improve the corrosion resistance of high-strength steel in marine converter stations. SUMMARY

[0004] To solve the above technical problems, the present application provides a structure optimization method and system for a marine converter station based on simulation, which improves the safety and reliability of the operation of the marine converter station.

[0005] In a first aspect, the present application provides a structure optimization method for a marine converter station based on simulation, comprising:

[0006] According to the service environment data of the target converter station, corrosion simulation of the target converter station is performed through a virtual simulation platform to obtain a simulation dataset under the service environment;

[0007] According to the simulation dataset, the interaction effect of the temperature period gradient and the stress corrosion coefficient is analyzed through a preset multi-field coupled analysis model, and then the dominant influencing factor affecting the corrosion rate is determined;

[0008] According to the main influencing factor and the simulation data set, a time series analysis method is used to track the dynamic change characteristics of the pit depth threshold in the corrosion process, and distribution rule data of the pit depth threshold under the tidal period change are obtained;

[0009] According to the distribution rule data, a failure probability distribution map of the target converter station is generated;

[0010] According to the failure probability distribution map, a dynamic protection parameter set based on the stress corrosion coefficient and the fatigue corrosion factor is generated, and the structure of the target converter station is optimized based on the dynamic protection parameter set.

[0011] The embodiment of the application provides a structure optimization method of a marine converter station based on simulation. Firstly, through a high-precision virtual simulation platform, the corrosion behavior of a target converter station under its real service environment is simulated in a full process and dynamically, so that the limitations of traditional single and static corrosion prevention measures are overcome. Then, by introducing a multi-field coupling analysis model, the complex interaction effects between key physical fields such as temperature and stress are scientifically analyzed, and the key influencing factors that truly dominate the corrosion process under different space-time scales are accurately identified, rather than relying on experience. Further, the time series analysis method is applied to the dynamic tracking of the pit depth threshold, the evolution rule of the corrosion damage under the action of periodic environmental forces such as tides is revealed, and an intuitive failure probability distribution map is generated, finally realizing the targeted and fine optimization of the structure of the converter station, and building a complete technical closed loop from "virtual environment simulation" to "physical structure optimization", which significantly improves the safety and reliability of the marine converter station in the harsh marine environment.

[0012] In a possible implementation manner, according to the service environment data of the target converter station, the target converter station is simulated for corrosion by a virtual simulation platform, and a corrosion characteristic data set under the service environment is obtained, including:

[0013] The tidal period change data, the salinity fluctuation characteristic data, the dissolved oxygen content distribution data, the temperature period gradient and the marine biological adhesion data of the target converter station are obtained, and the service environment data is constructed;

[0014] According to the service environment data, a multi-field coupling model of a preset material under the service environment is constructed by a virtual simulation platform, and an initial environment parameter set is obtained;

[0015] By using a finite element analysis method, according to the initial environment parameter set, the stress distribution and the electrochemical reaction of the preset material under the multi-field coupling environment are simulated, and a surface corrosion distribution map is obtained;

[0016] According to the surface corrosion distribution map, the initial environmental parameter set is adjusted through an iterative optimization algorithm to obtain an optimized corrosion rate data set and a pit depth threshold data set;

[0017] The corrosion characteristic data set is obtained according to the corrosion rate data set and the pit depth threshold data set;

[0018] The service environment data and the corrosion characteristic data set are spatiotemporally aligned to obtain the corrosion characteristic data set.

[0019] The embodiment of the application provides a method for obtaining a corrosion characteristic data set, which systematically integrates multi-source heterogeneous environment data such as a tidal period, a salinity fluctuation, a dissolved oxygen content, a temperature gradient and biological adhesion, and ensures completeness and authenticity of simulation input conditions. By constructing a multi-field coupling model and using finite element analysis, the stress distribution and the electrochemical reaction process of the material in the actual service environment can be accurately simulated, so that a high-fidelity surface corrosion distribution is obtained. In addition, the iterative optimization algorithm is introduced in the embodiment of the application, which can dynamically adjust the environmental parameters according to the preliminary simulation results, so that the output corrosion rate and pit depth data set continuously approach the real situation, greatly improving the accuracy and reliability of the data set. Finally, through the spatiotemporal alignment technology, the environment data and the corrosion characteristic data are deeply fused, ensuring the consistency of the data in the time dimension and the space dimension, and laying a high-quality data cornerstone for subsequent all data analysis and structure optimization.

[0020] In a possible implementation manner, according to the simulation simulation data set, the interaction effect of the temperature period gradient and the stress corrosion coefficient is analyzed through a preset multi-field coupling analysis model, and then the dominant influence factor affecting the corrosion rate is determined, including:

[0021] The temperature period gradient and the stress corrosion coefficient of the service environment are determined according to the simulation simulation data set;

[0022] The interaction between the temperature period gradient and the stress corrosion coefficient is analyzed through the multi-field coupling analysis model, and a basic mapping relationship between different temperatures and different stress corrosion coefficients is obtained;

[0023] According to the basic mapping relationship and the simulation simulation data set, a correlation mapping of each influence factor and the corrosion rate is constructed, and then the influence ranking of each influence factor is determined, the influence factor including a wave impact frequency, a biological adhesion period, a temperature period gradient, a stress corrosion coefficient and a salinity fluctuation amplitude;

[0024] According to the influence ranking of each influence factor and the preset number of dominant influence factors, a plurality of dominant influence factors are determined.

[0025] The embodiment of the application provides a method for determining dominant influence factors, which deeply analyzes the interaction of two core parameters, temperature period gradient and stress corrosion coefficient, through a multi-field coupling model, and establishes a basic mapping relationship between them and a corrosion rate, thereby deepening the understanding of the corrosion behavior from a mechanism level. On this basis, the embodiment performs systematic correlation mapping and quantitative influence ranking on all candidate influence factors, so as to filter out a number of dominant influence factors with the most concise quantity but the most critical contribution according to a preset scientific criterion. This process effectively eliminates the interference of secondary factors, so that the subsequent protection resources can be concentrated on the most critical control variables, greatly improving the efficiency and accuracy of the protection measures, and thereby improving the safety and reliability of the offshore converter station operation.

[0026] In a possible implementation manner, the time series analysis method is used to track the dynamic change characteristics of the pit depth threshold in the corrosion process according to the dominant influence factors and the simulation simulation data set, to obtain distribution rule data of the pit depth threshold under the tidal period change, including:

[0027] Each sub-data set corresponding to each of the dominant influence factors is extracted from the simulation simulation data set;

[0028] The time series analysis method is used to preliminarily decompose the change of the pit depth threshold according to the simulation simulation data set, to obtain the time sequence distribution of the dynamic change characteristics;

[0029] An initial periodic force influence model is constructed according to the tidal period change data in the simulation simulation data set and the time sequence distribution of the dynamic change characteristics;

[0030] The correlation characteristics of the periodic force on the change of the pit depth threshold are extracted from the simulation simulation data set according to the initial periodic force influence model, to obtain initial distribution rule data of the pit depth threshold under the tidal period change;

[0031] The support vector machine method is used to determine the influence coefficient of each of the dominant influence factors on the change of the pit depth threshold according to each of the sub-data sets and the initial distribution rule data;

[0032] The initial distribution rule data is corrected according to each of the influence coefficients and each of the sub-data sets, to obtain the distribution rule data of the pit depth threshold under the tidal period change.

[0033] The embodiment of the application provides a method for obtaining pit depth distribution regularity data, first, a time series analysis method is used, which can effectively peel off the dynamic change characteristics of pit depth with time from massive simulation data, especially capture the response regularity under the influence of tidal period change, and obtain the time sequence distribution of the dynamic change characteristics. Further, an initial periodic force influence model is constructed by combining the simulation data set, so as to capture the influence of the macro environment period. Finally, a support vector machine (SVM) is used to quantitatively evaluate the role of each dominant influence factor in the change of pit depth, determine the influence coefficient of each factor, and correct and optimize the initial distribution regularity by using the coefficients, so as to obtain the distribution regularity data of the pit depth threshold under the influence of the tidal period change. The combination strategy of'model preliminary screening + data driven correction' significantly enhances the capture accuracy and description ability of the complex nonlinear corrosion process evolution regularity, guarantees the rationality and effectiveness of subsequent structure optimization, and further improves the safety and reliability of the offshore converter station operation.

[0034] In a possible implementation manner, the generating, according to the distribution regularity data, of a failure probability distribution map of the target converter station comprises:

[0035] According to the distribution regularity data, the salinity fluctuation characteristic data and the dissolved oxygen content distribution data in the simulation simulation data set, a mapping relationship model of the pit depth threshold and the marine corrosion period is constructed;

[0036] According to the mapping relationship model, the simulation simulation data set and the operation environment parameters of the target converter station, a k-means clustering algorithm is used to determine the corrosion degree of each part of the target converter station;

[0037] According to the corrosion degree of each part of the target converter station, a corrosion rate data set in the simulation simulation data set is weighted and updated, to obtain the corrosion rate distribution data of each part of the target converter station;

[0038] Through a Monte Carlo simulation method, in combination with the corrosion rate distribution data and each dominant influence factor, failure probability distribution data of each part of the target converter station is generated;

[0039] Through a visualization tool, the failure probability distribution data is drawn in the form of a two-dimensional curve, to obtain the failure probability distribution map.

[0040] The embodiment of the application provides a failure probability distribution graph generation method, which closely combines the physical corrosion process and the probability risk assessment by constructing a mapping relationship model between the pit depth threshold and the marine corrosion period. The K-means clustering algorithm is used to intelligently identify and classify the corrosion degree of different parts of the converter station. Then, the Monte Carlo simulation, a powerful probability analysis tool, is introduced to calculate the failure probability of each part under the influence of uncertain environmental factors through a large number of random sampling. Finally, a visual failure probability distribution graph is generated through visualization technology. The failure probability distribution graph can intuitively show the high-risk areas and weak links in the entire converter station structure, and provide an extremely intuitive and reliable decision basis for subsequent development of differentiated and key protection strategies, thereby improving the safety and reliability of the offshore converter station operation.

[0041] Further, the mapping relationship model between the pit depth threshold and the marine corrosion period is constructed according to the distribution law data, the salinity fluctuation characteristic data and the dissolved oxygen content distribution data in the simulation data set, and includes:

[0042] The distribution law data is subjected to time series decomposition to obtain a periodic change sequence;

[0043] According to the salinity fluctuation characteristic data, the dissolved oxygen content distribution data and the periodic change sequence, a time series prediction is performed using an autoregressive moving average model to obtain a predicted period sequence, wherein the autoregressive moving average model is a linear combination of an autoregressive term and a moving average term based on historical data, and is constructed by determining the model parameters through maximum likelihood estimation;

[0044] According to the predicted period sequence and the simulation data set, a non-linear relationship between the pit depth threshold and the corrosion period is fitted by a polynomial function, and then the mapping relationship model between the pit depth threshold and the marine corrosion period is constructed.

[0045] In the construction of the key mapping relationship model, the embodiment of the application adopts more precise and reliable technical means. By time series decomposition of the distribution law data, the internal periodic change sequence can be obtained. Then, combined with key environmental parameters such as salinity and dissolved oxygen, a high-precision time series prediction is performed using an autoregressive moving average (ARMA) model to obtain the future corrosion period sequence. On this basis, the non-linear relationship between the pit depth threshold and the corrosion period can be accurately described by polynomial function fitting, so that the established mapping relationship model not only has excellent prediction ability, but also can deeply reveal the dynamic nature of the corrosion process, thereby pushing the corrosion risk assessment from qualitative to quantitative, greatly improving the credibility of the evaluation results, improving the accuracy and efficiency of subsequent structure optimization of the target converter station, and thereby improving the safety and reliability of the offshore converter station operation.

[0046] In a possible implementation, the generating, according to the failure probability distribution diagram, of a dynamic protection parameter set based on a stress corrosion coefficient and a fatigue corrosion factor includes:

[0047] In combination with the failure probability distribution diagram, a plurality of rounds of numerical simulation is performed on the target converter station under the service environment to obtain a numerical simulation data set and a nonlinear characteristic sequence of salt concentration fluctuation;

[0048] A preset medium corrosion index and a preset electrochemical reaction parameter are adjusted according to the nonlinear characteristic sequence of salt concentration fluctuation to obtain a corrosion parameter set, wherein the preset medium corrosion index and the preset electrochemical reaction parameter are determined based on a plurality of corrosion experiments under a marine environment;

[0049] The numerical simulation data set and the corrosion parameter set are estimated, a regression analysis is used to construct a quantitative relationship expression between a pit depth threshold and a marine corrosion period, and then a corrosion period model is determined;

[0050] Based on the corrosion period model, a genetic algorithm is used to optimize an initial stress corrosion coefficient and an initial fatigue corrosion factor to obtain the dynamic protection parameter set, wherein the initial stress corrosion coefficient and the initial fatigue corrosion factor are determined according to the numerical simulation data set, and the dynamic protection parameter set includes material strength distribution data, material thickness distribution data, and connection stiffness distribution data.

[0051] The embodiment of the application provides a method for generating a dynamic protection parameter set. The method can capture the nonlinear change characteristics of environmental factors such as salt concentration fluctuation by performing a plurality of rounds of numerical simulation under a real service environment, and dynamically adjust the medium corrosion index and the electrochemical reaction parameter accordingly, so that the model parameters can closely match the dynamic changes of the actual environment. In particular, by using a genetic algorithm to optimize the initial stress corrosion coefficient and the fatigue corrosion factor, an optimal dynamic protection parameter set (including material strength, thickness, and connection stiffness distribution) can be automatically found, which provides a core parameter basis for subsequent realization of real "dynamic protection" and "on-demand optimization", and ensures that the protection scheme is always in an optimal or near-optimal state. The embodiment of the application has high adaptability and optimization capability in the final stage of generating a dynamic protection parameter set and performing structure optimization.

[0052] In a possible implementation, the structure optimization of the target converter station based on the dynamic protection parameter set includes:

[0053] An initial structure parameter matrix is constructed according to the dynamic protection parameter set;

[0054] The initial structure parameter matrix is iteratively updated several times by using the finite element analysis method, and a final optimized structure parameter matrix is obtained, wherein in each iteration process, a current structure finite element grid model corresponding to the current structure parameter matrix is generated by using the finite element analysis method according to the current structure parameter matrix, and the current stress distribution and the current deformation are calculated to obtain current structure response data; the material strength distribution data and the material thickness distribution data in the current structure parameter matrix are adjusted according to the current structure response data, an updated structure parameter matrix is obtained, and the next iteration update of the updated structure parameter matrix is performed;

[0055] A structure adjustment scheme is generated according to the optimized structure parameter matrix, and the structure adjustment scheme includes thickness adjustment values and connection stiffness adjustment values of specific components;

[0056] The target converter station is structurally optimized according to the structure adjustment scheme.

[0057] The embodiments of the present application provide a method for structurally optimizing a target converter station, by constructing an initial structure parameter matrix and iteratively updating the initial structure parameter matrix several times by using the finite element analysis method. Each iteration is based on current structure response data to adjust the structure parameters, such as changing the material thickness and the connection stiffness. This iterative optimization process ensures that the final optimized structure parameter matrix can fully meet the safety threshold in the expected corrosion environment, so that the finally generated structure adjustment scheme can directly guide engineering practice, and ensure that the finally optimized structure reaches the best balance among strength, stiffness and fatigue resistance, thereby improving the safety and reliability of the offshore converter station operation.

[0058] In a second aspect, the embodiments of the present application provide a structure optimization system for an offshore converter station based on simulation, which includes a simulation module, a dominant influence factor determination module, a data analysis module, a distribution map generation module and a structure optimization module;

[0059] The simulation module is configured to perform corrosion simulation of the target converter station through a virtual simulation platform according to service environment data of the target converter station, and obtain a simulation data set in a service environment.

[0060] The dominant influence factor determination module is configured to analyze the interaction effect of the temperature period gradient and the stress corrosion coefficient by using a preset multi-field coupling analysis model according to the simulation data set, and determine a dominant influence factor affecting the corrosion rate.

[0061] The data analysis module is configured to track the dynamic change characteristics of the pit depth threshold in the corrosion process by using a time series analysis method according to the dominant influence factor and the simulation data set, and obtain distribution rule data of the pit depth threshold under the tidal period change.

[0062] The distribution map generation module is configured to generate a failure probability distribution map of the target converter station according to the distribution law data.

[0063] The structure optimization module is configured to generate a dynamic protection parameter set based on a stress corrosion coefficient and a fatigue corrosion factor according to the failure probability distribution map, and perform structure optimization on the target converter station based on the dynamic protection parameter set.

[0064] Further, the simulation module performs corrosion simulation and modeling on the target converter station according to service environment data of the target converter station through a virtual simulation platform, to obtain a corrosion characteristic data set under a service environment, including:

[0065] Obtain tidal cycle change data, salinity fluctuation characteristic data, dissolved oxygen content distribution data, temperature cycle gradient, and marine biological attachment data of the target converter station, and construct the service environment data;

[0066] According to the service environment data, a multi-field coupling model of a preset material under a service environment is constructed through a virtual simulation platform, to obtain an initial environment parameter set;

[0067] According to the initial environment parameter set, a stress distribution and an electrochemical reaction of the preset material under a multi-field coupling environment are simulated by using a finite element analysis method, to obtain a surface corrosion distribution map;

[0068] According to the surface corrosion distribution map, the initial environment parameter set is adjusted through an iterative optimization algorithm, to obtain an optimized corrosion rate data set and a pit depth threshold data set;

[0069] The corrosion characteristic data set is constructed according to the corrosion rate data set and the pit depth threshold data set;

[0070] The service environment data and the corrosion characteristic data set are spatio-temporally aligned, to construct the corrosion characteristic data set. BRIEF DESCRIPTION OF DRAWINGS

[0071] Figure 1 A flowchart of a structure optimization method of a marine converter station based on simulation and modeling provided by an embodiment of the present application;

[0072] Figure 2 A structure diagram of a structure optimization system of a marine converter station based on simulation and modeling provided by an embodiment of the present application. DETAILED DESCRIPTION

[0073] With reference to the drawings and the embodiments of the present application, the technical solutions in the embodiments of the present application will be described clearly and completely. Obviously, the described embodiments are only a part of the embodiments of the present application, but not all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of the present application.

[0074] It should be noted that the step numbers in the text are only for the convenience of the explanation of the specific embodiments, and do not serve as the function of limiting the execution sequence of the steps. In the description of the present application, the terms "first", "second" are only for the purpose of description, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features limited by "first", "second" can explicitly or implicitly include one or more of the features.

[0075] Embodiment one:

[0076] As shown in Figure 1 Embodiment one provides a structure optimization method of a marine converter station based on simulation, comprising steps S1-S5:

[0077] Step S1, according to the service environment data of the target converter station, corrosion simulation of the target converter station is carried out through a virtual simulation platform, and a simulation data set under the service environment is obtained;

[0078] Step S2, according to the simulation data set, the interaction effect of temperature cycle gradient and stress corrosion coefficient is analyzed through a preset multi-field coupling analysis model, and then the dominant influence factor affecting the corrosion rate is determined;

[0079] Step S3, according to the dominant influence factor and the simulation data set, the dynamic change characteristics of the pit depth threshold value in the corrosion process are tracked by using a time series analysis method, and distribution rule data of the pit depth threshold value under the tidal cycle change are obtained;

[0080] Step S4, according to the distribution rule data, a failure probability distribution map of the target converter station is generated;

[0081] Step S5, according to the failure probability distribution map, a dynamic protection parameter set based on the stress corrosion coefficient and the fatigue corrosion factor is generated, and the structure of the target converter station is optimized based on the dynamic protection parameter set.

[0082] The embodiment of the application provides a structure optimization method for a marine converter station based on simulation and analog, first, through a high-precision virtual simulation platform, a full-process and dynamic digital simulation is performed on the corrosion behavior of a target converter station under a real service environment of the target converter station, so that the limitation of a traditional single and static corrosion prevention measure is overcome. Then, by introducing a multi-field coupling analysis model, complex interaction effects between key physical fields such as temperature and stress are scientifically analyzed, and key influence factors that truly dominate the corrosion process under different space-time scales are accurately identified, instead of relying on experience. Further, a time series analysis method is applied to dynamic tracking of an erosion pit depth threshold, evolution rules of corrosion damage under the action of periodic environmental forces such as tides are revealed, and based on this, an intuitive failure probability distribution diagram is generated, finally, targeted and refined optimization of the converter station structure is realized, a complete technical closed loop from “virtual environment simulation” to “physical structure optimization” is constructed, and the safety and reliability of the marine converter station in long-term service in a harsh marine environment are significantly improved.

[0083] In a possible implementation manner, in step S1, according to service environment data of a target converter station, corrosion simulation and analog of the target converter station is performed through a virtual simulation platform, and corrosion characteristic data sets under a service environment are obtained, including:

[0084] Tidal cycle change data, salinity fluctuation characteristic data, dissolved oxygen content distribution data, temperature cycle gradient and marine biological adhesion data of the target converter station are obtained, and the service environment data is constructed;

[0085] According to the service environment data, a multi-field coupling model of a preset material under a service environment is constructed through a virtual simulation platform, and an initial environment parameter set is obtained;

[0086] By using a finite element analysis method, stress distribution and electrochemical reaction of the preset material under a multi-field coupling environment are simulated according to the initial environment parameter set, and a surface corrosion distribution map is obtained;

[0087] According to the surface corrosion distribution map, the initial environment parameter set is adjusted through an iterative optimization algorithm, and an optimized corrosion rate data set and an erosion pit depth threshold data set are obtained;

[0088] The corrosion characteristic data sets are obtained according to the corrosion rate data set and the erosion pit depth threshold data set;

[0089] The service environment data and the corrosion characteristic data sets are spatio-temporally aligned, and the corrosion characteristic data sets are constructed.

[0090] This application provides a method for obtaining a corrosion characteristic dataset, systematically integrating multi-source heterogeneous environmental data such as tidal cycles, salinity fluctuations, dissolved oxygen content, temperature gradients, and biofouling, ensuring the completeness and realism of the simulation input conditions. By constructing a multi-field coupled model and employing finite element analysis, the stress distribution and electrochemical reaction processes of materials under actual service environments can be accurately simulated, thereby obtaining high-fidelity surface corrosion distribution. Furthermore, this application introduces an iterative optimization algorithm, which dynamically adjusts environmental parameters based on preliminary simulation results, continuously approximating the output corrosion rate and pit depth datasets to real-world conditions, significantly improving the accuracy and reliability of the dataset. Finally, through spatiotemporal alignment technology, environmental data and corrosion characteristic data are deeply fused, ensuring data consistency in both temporal and spatial dimensions, laying a high-quality data foundation for all subsequent data analysis and structural optimization.

[0091] In a preferred embodiment, the service environment of the high-strength steel material of the offshore converter station is constructed using a virtual simulation platform, and service environment data is obtained through simulation. First, based on tidal cycle variations, a sine function is used to simulate tidal height changes, with a period of 12.42 hours and an amplitude of 2.5 meters. Based on NOAA tidal data, Fourier transform analysis is performed using the SciPy library in Python to extract tidal frequency components, ensuring a simulation accuracy of 0.01 meters. Salinity fluctuation characteristics are analyzed using measured data collected from the southern North Sea, with a salinity range of 30-35 psu and a fluctuation frequency of twice daily. A random walk algorithm is used to generate time series data, with the standard deviation controlled within 0.5 psu. The normality of the fluctuation distribution is verified using the statistical toolbox of Matlab. Dissolved oxygen distribution was analyzed using a three-dimensional finite element method. A 1000m × 1000m × 50m marine model was built using COMSOL Multiphysics software, with surface dissolved oxygen set at 8 mg / L and deep dissolved oxygen at 5 mg / L. The dissolved oxygen gradient was calculated using the convection-diffusion equation, with the error controlled within 0.1 mg / L. Temperature periodic gradients were simulated using an annual periodic sinusoidal model, with a temperature range of 5-25℃ and a daily variation of 2℃. Heat conduction was simulated using ANSYS Fluent with a mesh resolution of 0.1m, and the calculated temperature field convergence residual was less than 10^-5. Marine biofilm attachment was simulated using a Poisson distribution, assuming an attachment density of 50 organisms per square meter. The Monte Carlo method was used to generate the attachment point distribution, and with a biofilm coverage of 0.3, the corrosion-promoting coefficient was calculated to be 1.2. Then, an electrochemical corrosion model was adopted, based on the Tafel equation, setting the corrosion potential of high-strength steel to -0.65V and the exchange current density to 10^-6 A / cm^2. The initial corrosion rate constant was calculated to be 0.15 mm / year using the finite element method. The pit depth threshold was fitted using the Gumbel extreme value distribution, yielding an average depth of 0.2 mm and a standard deviation of 0.05 mm. All data were integrated using a Python script to generate a corrosion characteristic dataset.

[0092] In one possible implementation, in step S2, the interaction effect between the temperature periodic gradient and the stress corrosion coefficient is analyzed using a preset multi-field coupling analysis model based on the simulation dataset, thereby determining the dominant influencing factors affecting the corrosion rate, including:

[0093] The temperature cycle gradient and stress corrosion coefficient of the service environment are determined based on the simulation dataset.

[0094] The interaction between the temperature periodic gradient and the stress corrosion coefficient is analyzed by the multi-field coupling analysis model to obtain the basic mapping relationship between different temperatures and different stress corrosion coefficients.

[0095] Based on the basic mapping relationship and the simulation dataset, the correlation mapping between each influencing factor and the corrosion rate is constructed respectively, and then the influence ranking of each influencing factor is determined. The influencing factors include wave impact frequency, biological attachment cycle, temperature cycle gradient, stress corrosion coefficient and salinity fluctuation amplitude.

[0096] Based on the ranking of the influence of each of the aforementioned influencing factors and the preset number of dominant influencing factors, several dominant influencing factors are determined.

[0097] This application provides a method for determining dominant influencing factors. It utilizes a multi-field coupling model to deeply analyze the interaction between two core parameters: temperature periodic gradient and stress corrosion coefficient, establishing a fundamental mapping relationship between them and corrosion rate, thus deepening the understanding of corrosion behavior at the mechanistic level. Based on this, this embodiment systematically correlates and maps all candidate influencing factors and quantifies their influence ranking. This allows for the selection of a few dominant influencing factors with the most concise yet crucial contribution, according to preset scientific criteria. This process effectively eliminates the interference of secondary factors, enabling subsequent protective resources to be concentrated on the most critical control variables, greatly improving the efficiency and accuracy of protective measures, and ultimately enhancing the safety and reliability of offshore converter station operations.

[0098] In a preferred embodiment, the corrosion rate dataset in the simulation dataset is preprocessed by removing outliers using a data cleaning method, resulting in a cleaned corrosion dataset. This cleaned dataset contains the initial value distributions of temperature cycle parameters and stress corrosion parameters. Principal component analysis (PCA) is applied to extract key feature vectors from the cleaned dataset. PCA involves decomposing the covariance matrix to obtain orthogonal basis vectors representing the directions of maximum data variance, resulting in an extracted feature vector set. This feature vector set represents the low-dimensional representation of temperature cycle and stress corrosion. Subsets of temperature cycle parameters and stress corrosion parameters are then obtained from the extracted feature vector set. A multi-field coupled analysis model is constructed based on these subsets. This model integrates the temperature and stress fields into a finite element simulation framework. Through mesh generation and boundary condition settings, a simulation interaction process is established, yielding the model's simulation output, which includes the preliminary field distribution of interactive effects. Based on the model simulation output, the K-nearest neighbor algorithm is applied to classify the interactive influence regions. The K-nearest neighbor algorithm refers to determining the category based on the majority vote of the nearest neighbor samples calculated by Euclidean distance, and obtaining a set of classified interactive regions. The set of classified interactive regions is divided into high and low interaction intensity zones, which is the basic mapping relationship between different temperatures and different stress corrosion coefficients.

[0099] Then, based on the fundamental mapping relationship and the simulation dataset, the correlation mapping between each influencing factor and the corrosion rate is constructed to determine the ranking of the influence of each influencing factor. Taking wave impact frequency and bioattachment cycle as examples, a support vector machine model can be used for classification processing when constructing the correlation mapping between frequency influence and cycle effect. Wave impact frequency may exhibit a high-frequency characteristic of multiple times per day, while bioattachment shows a longer periodicity, such as a significant increase once a month. Assuming a wave impact frequency of 10 times per day and a bioattachment cycle of 30 days, through classification processing, wave impact can be initially ranked as the dominant factor, with a direct influence weight of approximately 0.6 on the corrosion rate, while the bioattachment weight is 0.4. In addition, a comprehensive influence matrix can be constructed to represent the correlation mapping between each influencing factor and the corrosion rate and pit depth. Through matrix decomposition, the evolution trend of corrosion rate and pit depth is decomposed into multiple independent factors. Assuming that in the final dominant factor combination, wave impact and temperature cycle account for 0.4 and 0.3 respectively, this combination can more comprehensively reflect the degradation mechanism of materials in complex environments, providing an important reference for subsequent protection design.

[0100] In one possible implementation, in step S3, the step of tracking the dynamic change characteristics of the pit depth threshold during corrosion using time series analysis based on the dominant influencing factor and the simulation dataset, to obtain the distribution data of the pit depth threshold under tidal cycle variations, includes:

[0101] Extract the sub-datasets corresponding to each of the dominant influencing factors from the simulation dataset;

[0102] Based on the simulation dataset, the time series analysis method is used to preliminarily decompose the changes in the pit depth threshold and obtain the time series distribution of dynamic change characteristics.

[0103] Based on the tidal cycle variation data and the temporal distribution of the dynamic variation characteristics in the simulation dataset, an initial periodic force influence model is constructed.

[0104] Based on the initial periodic force influence model, the correlation features between periodic force and pit depth threshold change are extracted from the simulation dataset to obtain the initial distribution law data of pit depth threshold under tidal cycle change.

[0105] Based on each of the aforementioned subsets and the initial distribution pattern data, the support vector machine method is used to determine the influence coefficients of each of the aforementioned dominant influencing factors on the change of pit depth threshold.

[0106] The initial distribution data is corrected based on the respective influence coefficients and the respective subsets to obtain the distribution data of the pit depth threshold under tidal cycle changes.

[0107] This application provides a method for obtaining data on the distribution patterns of corrosion pit depth. First, time series analysis is used to effectively extract the dynamic variation characteristics of corrosion pit depth over time from massive amounts of simulation data, particularly capturing its response patterns under tidal cycles, thus obtaining the time-series distribution of dynamic variation characteristics. Furthermore, an initial periodic force influence model is constructed using the simulation dataset to capture the impact of macroscopic environmental cycles. Finally, Support Vector Machine (SVM), an advanced machine learning method, is used to quantify and evaluate the roles of each dominant influencing factor in corrosion pit depth variation, determine their respective influence coefficients, and use these coefficients to correct and optimize the initial distribution pattern, obtaining the distribution pattern data of corrosion pit depth thresholds under tidal cycles. This combined strategy of "model initial screening + data-driven correction" significantly enhances the accuracy and descriptive ability of capturing the evolution patterns of complex nonlinear corrosion processes, ensuring the rationality and effectiveness of subsequent structural optimization, thereby improving the safety and reliability of offshore converter station operation.

[0108] In a preferred embodiment, the dominant influencing factors are wave impact frequency and biofouling cycle. Therefore, relevant data on wave impact frequency and biofouling cycle are extracted from the simulation dataset to construct wave impact frequency dataset and biofouling cycle dataset. Then, basic parameters related to corrosion process tracking are obtained from the wave impact frequency dataset and biofouling cycle dataset, and the changing trend of pit depth can be decomposed using time series analysis. Assuming a certain marine platform environment, the daily wave impact frequency varies from 5 to 15 times, and the biofouling cycle is 20 to 40 days. Time series analysis can employ an autoregressive model to extract the periodic fluctuation characteristics of pit depth over time. For example, the analysis results show a positive correlation between the daily peak value of wave impact frequency and the increase in pit depth; for every two increases in frequency, the pit depth may increase by 0.03 mm. This decomposition helps identify the dynamic temporal distribution, providing data support for subsequent modeling. Furthermore, when constructing a periodic force influence model for data on tidal cycle changes, the frequency component of the tidal cycle can be extracted using the Fourier transform method. Assuming a tidal cycle of 12 hours, model analysis shows that the weight of tidal-induced periodic stress changes on pit depth is approximately 0.35. Through feature extraction, the model reveals an 80% peak synchronization rate between the tidal cycle and pit depth changes, indicating that tidal action is a significant driver of corrosion behavior. Initial distribution data of pit depth threshold under tidal cycle variations are obtained using this method. Then, a support vector machine (SVM) method is used to classify and rank the dominant influencing factors of wave impact frequency and biofouling cycle, dividing the dataset into high-frequency and low-frequency impacts, while also considering the length of the biofouling cycle. Assuming the classification results show a weight of 0.55 for wave impact frequency and 0.45 for the biofouling cycle, the priority distribution indicates that high-frequency impact has a more significant impact on the corrosion rate, thus determining the influence coefficients of each dominant influencing factor on pit depth threshold changes. The classification process utilizes kernel function optimization to differentiate the contributions of different factors to corrosion behavior, providing a basis for accurate assessment.

[0109] In one possible implementation, step S4, generating the failure probability distribution map of the target converter station based on the distribution data, includes:

[0110] Based on the distribution pattern data, the salinity fluctuation characteristic data and dissolved oxygen content distribution data in the simulation dataset, a mapping relationship model between the pit depth threshold and the marine corrosion cycle is constructed.

[0111] Based on the mapping relationship model, the simulation dataset, and the operating environment parameters of the target converter station, the degree of corrosion of each part of the target converter station is determined using the k-means clustering algorithm.

[0112] The corrosion rate dataset in the simulation dataset is weighted and updated according to the corrosion degree of each part of the target converter station to obtain the corrosion rate distribution data of each part of the target converter station.

[0113] By using the Monte Carlo simulation method, combined with the corrosion rate distribution data and each of the dominant influencing factors, failure probability distribution data for each part of the target converter station are generated.

[0114] The failure probability distribution data is plotted as a two-dimensional curve using visualization tools to obtain the failure probability distribution map.

[0115] This application provides a method for generating a failure probability distribution map. By constructing a mapping model between pit depth thresholds and marine corrosion cycles, the physical corrosion process is closely integrated with probabilistic risk assessment. K-means clustering is used to intelligently identify and classify the corrosion degree of different parts of the converter station. Furthermore, the method introduces Monte Carlo simulation, a powerful probabilistic analysis tool, to calculate the failure probability of each part under the influence of uncertain environmental factors through extensive random sampling. Finally, visualization technology is used to generate a clear and concise failure probability distribution map. This failure probability distribution map can intuitively display the high-risk areas and weak points in the entire converter station structure, providing a highly intuitive and reliable decision-making basis for subsequently developing differentiated and focused protection strategies, thereby improving the safety and reliability of offshore converter station operations.

[0116] In a preferred embodiment, based on the mapping relationship model and the simulation dataset, and combined with the operating environment parameters of the converter station, the k-means clustering algorithm is used to classify the weak links of the converter station, obtaining a distribution set of weak links. If the corrosion rate constant of a certain type of link in the distribution set of weak links exceeds a preset threshold, the time-series data of that type of link is weighted to obtain weighted corrosion rate distribution data. Using the Monte Carlo simulation method, combined with the weighted corrosion rate and various dominant influencing factors, failure probability distribution data for each link of the converter station is generated. Monte Carlo simulation refers to calculating the failure probability of each link under different environmental conditions through multiple random sampling, based on a probability model of weighted corrosion rate and dominant influencing factors, to obtain failure probability distribution data. Based on the failure probability distribution data, a priority sorting algorithm is used to determine the high-risk distribution sequence. The priority sorting algorithm uses the failure probability value as the sorting basis and uses a quick sorting method to arrange the links of the converter station in descending order to obtain the high-risk distribution sequence, where the links at the top of the sequence indicate a higher failure probability. The set of weak links is extracted from the high-risk distribution sequence. The weak link set refers to the set of links whose failure probability exceeds a preset threshold. If the failure probability of a certain link is greater than the preset threshold, it is included in the weak link set, and the distribution data of the weak links is obtained. Based on the weak link set and the pit depth threshold, a threshold mapping model is constructed. The threshold mapping model refers to fitting a mapping function using the least squares method through the nonlinear relationship between pit depth and failure probability, where the pit depth threshold refers to the critical corrosion depth that leads to link failure, and the mapping model parameters are obtained. The failure probability distribution is finely adjusted using a linear interpolation method. The linear interpolation method refers to inserting new probability values ​​between failure probability distribution data points based on the mapping model parameters to generate a continuous failure probability distribution curve, and obtain the adjusted failure probability distribution data. Based on the adjusted failure probability distribution data, a final failure probability distribution map is generated. Generating the failure probability distribution map means using visualization tools to plot the adjusted failure probability distribution data as a two-dimensional curve, where the horizontal axis represents the converter station links and the vertical axis represents the failure probability, resulting in the final failure probability distribution map.

[0117] Furthermore, the step of constructing a mapping model between the pit depth threshold and the marine corrosion cycle based on the distribution pattern data, the salinity fluctuation characteristic data and dissolved oxygen content distribution data in the simulation dataset includes:

[0118] The distribution data is decomposed into a time series to obtain a periodically changing sequence;

[0119] Based on the salinity fluctuation characteristic data, the dissolved oxygen content distribution data, and the periodic change sequence, an autoregressive moving average model is used for time series prediction to obtain the predicted periodic sequence. The autoregressive moving average model is a linear combination of autoregressive terms and moving average terms based on historical data, and is constructed after determining the model parameters through maximum likelihood estimation.

[0120] Based on the predicted cycle sequence and the simulation dataset, a mapping model between the pit depth threshold and the corrosion cycle is constructed by fitting the nonlinear relationship between the pit depth threshold and the corrosion cycle using a polynomial function.

[0121] In constructing the key mapping relationship model, this application employs more precise and reliable technical means. By performing time series decomposition on the distribution data, its inherent periodic variation sequence can be obtained. Then, combined with key environmental parameters such as salinity and dissolved oxygen, an autoregressive moving average (ARMA) model is used for high-precision time series prediction to obtain the future corrosion cycle sequence. Based on this, through polynomial function fitting, the nonlinear relationship that commonly exists between the pit depth threshold and the corrosion cycle can be accurately characterized. This makes the established mapping relationship model not only have excellent predictive ability but also profoundly reveal the dynamic nature of the corrosion process, thereby shifting the assessment of corrosion risk from qualitative to quantitative, significantly improving the credibility of the assessment results, increasing the accuracy and efficiency of subsequent structural optimization of the target converter station, and ultimately improving the safety and reliability of offshore converter station operation.

[0122] In a preferred embodiment, a periodic variation trend is extracted from the distribution data using a time series decomposition method to obtain a periodic variation sequence. This time series decomposition involves separating the trend, seasonality, and residual components of the time series and calculating the frequency components of the periodic variation sequence using a Fast Fourier Transform. Based on the periodic variation sequence, combined with salinity fluctuation characteristics and dissolved oxygen content distribution data, an autoregressive moving average model is used for time series prediction to obtain a predicted periodic sequence. The autoregressive moving average model is a linear combination of autoregressive and moving average terms based on historical data, and the model parameters are determined through maximum likelihood estimation. By simulating the dataset and the predicted periodic sequence, a mapping model between the pit depth threshold and the marine corrosion cycle is constructed. This mapping model involves fitting a nonlinear relationship between the pit depth threshold and the corrosion cycle using a polynomial function, and the model parameters are obtained through least squares fitting.

[0123] Specifically, after decomposing the distribution data into time series to obtain the periodic variation sequence, an autoregressive moving average model can be used for time series prediction. For example, in the marine environment, salinity fluctuations and dissolved oxygen content are key factors affecting corrosion rates. Assuming that salinity data for a certain sea area is collected monthly, with an annual fluctuation range of 32-35 PSU and dissolved oxygen content of 5-7 mg / L, analyzing the time series data of the past three years using an autoregressive moving average model predicts that salinity may fluctuate within 33±1 PSU and dissolved oxygen content may vary within the range of 6±0.5 mg / L in the next 12 months. This prediction provides a stable periodic variation sequence for subsequent mapping relationship modeling. When constructing the mapping relationship model between the pit depth threshold and the marine corrosion cycle, the least squares method is used to fit the parameters. For example, for the steel structure of a converter station, the pit depth threshold has a strong correlation with the tidal cycle. Assuming the tidal cycle is 12.4 hours, the pit depth threshold fluctuates between 0.5-1.2 mm. The corrosion rate constant was determined to be 0.02 mm / month using least squares fitting. The temperature influence factor showed that the corrosion rate increased by 0.001 mm / month for every 1°C increase in seawater temperature. This quantitative relationship provided a foundation for subsequent analysis. After obtaining the mapping model, and combining it with the converter station's operating environment parameters, such as seawater temperature of 20-25°C and flow velocity of 0.5-1.2 m / s, k-means clustering was used to classify weak points. For example, cable connection points and support piles in the converter station, due to long-term exposure to seawater, might be classified as high-corrosion-risk areas. The clustering results showed that the corrosion rate constant of the cable connection points was 0.025 mm / month, exceeding the preset threshold of 0.02 mm / month, requiring weighted processing. During weighted processing, the corrosion rate was assigned weights based on exposure time and flow velocity; for example, a weight of 1.2 was assigned for a flow velocity of 1 m / s, resulting in a weighted corrosion rate distribution. In one embodiment, the failure probability was calculated using Monte Carlo simulation. For example, after 10,000 simulations, the failure probability of cable connection points was found to be 15%, higher than the 10% for other parts. Based on this, the set of weak points was prioritized, and cable connection points were listed as high-risk points. A linear interpolation method was used to adjust the failure probability distribution, combining a mapping model of pit depth thresholds. For example, when the pit depth increased from 1.0 mm to 1.1 mm, the failure probability linearly interpolated from 15% to 16%, resulting in a more refined failure probability distribution map. This finer adjustment helps to accurately identify high-risk areas. For example, the predicted results of salinity fluctuations and dissolved oxygen content can be used to optimize the maintenance plan of the converter station. It should be noted that the failure probability distribution map combining Monte Carlo simulation with linear interpolation can provide a reliable basis for corrosion risk assessment for the long-term operation of the converter station. This multi-step analysis, from prediction to classification to probability calculation, forms a rigorous logical chain, which helps to improve the safety and durability of converter station equipment.

[0124] In one possible implementation, step S5, generating a dynamic protection parameter set based on the stress corrosion coefficient and fatigue corrosion factor according to the failure probability distribution map, includes:

[0125] Based on the failure probability distribution map, multiple rounds of numerical simulations were performed on the target converter station under the service environment to obtain the numerical simulation dataset and the nonlinear characteristic sequence of salinity fluctuations.

[0126] The preset medium corrosion index and preset electrochemical reaction parameters are adjusted according to the nonlinear characteristic sequence of the salinity fluctuation to obtain a set of corrosion parameters. The preset medium corrosion index and preset electrochemical reaction parameters are determined based on several corrosion experiments in a marine environment.

[0127] Estimate the numerical simulation dataset and the corrosion parameter set, and use regression analysis to construct a quantitative relationship expression between the pit depth threshold and the marine corrosion cycle, thereby determining the corrosion cycle model;

[0128] Based on the corrosion cycle model, the initial stress corrosion coefficient and the initial fatigue corrosion factor are optimized using a genetic algorithm to obtain the dynamic protection parameter set. The initial stress corrosion coefficient and the initial fatigue corrosion factor are determined based on the numerical simulation dataset. The dynamic protection parameter set includes material strength distribution data, material thickness distribution data, and connection stiffness distribution data.

[0129] This application provides a method for generating a dynamic protection parameter set. By conducting multiple rounds of numerical simulations under real service conditions, this method can capture the nonlinear variation characteristics of environmental factors such as salinity fluctuations and dynamically adjust the medium corrosion index and electrochemical reaction parameters accordingly, ensuring that the model parameters closely match the dynamic changes of the actual environment. In particular, by using a genetic algorithm to optimize the initial stress corrosion coefficient and fatigue corrosion factor, it can automatically find a set of optimal dynamic protection parameters (including material strength, thickness, and connection stiffness distribution). This provides the core parameter basis for subsequently achieving true "dynamic protection" and "on-demand optimization," ensuring that the protection scheme is always in an optimal or near-optimal state. This application demonstrates a high degree of adaptability and optimization capability in the final stage of generating the dynamic protection parameter set and performing structural optimization.

[0130] In a preferred embodiment, based on the failure probability distribution map, the nonlinear characteristic sequence of salinity fluctuations is analyzed through multiple rounds of numerical simulation. Salinity, as a key factor in marine corrosion, exhibits nonlinear fluctuations. Specifically, assuming salinity data for a certain sea area is collected quarterly, data from the past two years shows salinity fluctuating between 30-36 PSU, exhibiting seasonal nonlinear changes. Using the Fast Fourier Transform (FFT) method to decompose the time-series data and extract the nonlinear characteristic sequence, it was found that the maximum fluctuation amplitude reached 3 PSU, exceeding the preset threshold of 2 PSU. At this point, the media corrosion index and electrochemical reaction parameters need to be adjusted. The media corrosion index can be increased by 10% based on the salinity fluctuation amplitude, while the electrochemical reaction parameters are dynamically adjusted to a reaction rate constant of 0.015 mm / month, taking into account the historical trend of dissolved oxygen content changing from 5-8 mg / L. This adjustment forms a suitable set of corrosion parameters, providing a reliable foundation for subsequent modeling. Specifically, based on the corrosion parameter set, regression analysis is used to construct a quantitative relationship expression between the pit depth threshold and the marine corrosion cycle. Assuming the pitting depth threshold of a converter station's steel structure is 0.8-1.5 mm and the tidal cycle is 12.4 hours, polynomial regression analysis is used to determine the correlation between corrosion rate and salinity and temperature in the corrosion cycle model. For example, for every 1°C increase in seawater temperature, the corrosion rate increases by 0.002 mm / month, and for every 1 PSU increase in salinity, the corrosion rate increases by 0.001 mm / month. This quantitative expression clearly describes the impact of environmental factors on corrosion, laying the foundation for optimizing protection parameters. In one possible implementation, based on the corrosion cycle model, a genetic algorithm is used to optimize the stress corrosion coefficient and fatigue corrosion factor. Assuming the initial stress corrosion coefficient is 0.03 and the fatigue corrosion factor is 0.05, after 100 generations of iteration using the genetic algorithm, they are optimized to 0.028 and 0.045, respectively. If the optimized protection coefficient is lower than the preset threshold of 0.05, the material coating thickness or cathodic protection current density is further adjusted. For example, increasing the coating thickness to 2 mm increases the protection coefficient to 0.06, meeting the requirements. This dynamic set of protection parameters can effectively improve protection capabilities.

[0131] In one possible implementation, the structural optimization of the target converter station based on the dynamic protection parameter set includes:

[0132] Construct an initial structural parameter matrix based on the dynamic protection parameter set;

[0133] The initial structural parameter matrix is ​​iteratively updated several times using the finite element analysis method to obtain the final optimized structural parameter matrix. In each iteration, based on the current structural parameter matrix, a corresponding current structural finite element mesh model is generated using the finite element analysis method, and the current stress distribution and current deformation are calculated to obtain the current structural response data. Based on the current structural response data, the material strength distribution data and material thickness distribution data in the current structural parameter matrix are adjusted to obtain an updated structural parameter matrix, and the updated structural parameter matrix is ​​then iteratively updated again.

[0134] Based on the optimized structural parameter matrix, a structural adjustment scheme is generated, which includes the thickness adjustment value and connection stiffness adjustment value of specific components;

[0135] The target converter station is structurally optimized according to the structural adjustment scheme.

[0136] This application provides a method for structural optimization of a target converter station. An initial structural parameter matrix is ​​constructed, and finite element analysis is used for multiple iterative updates. Each iteration adjusts the structural parameters based on the current structural response data, such as changing material thickness and connection stiffness. This iterative optimization process ensures that the final optimized structural parameter matrix fully meets the safety threshold under the expected corrosive environment. This allows the final structural adjustment scheme to directly guide engineering practice, ensuring that the optimized structure achieves an optimal balance between strength, stiffness, and fatigue resistance, thereby improving the safety and reliability of the offshore converter station.

[0137] In a preferred embodiment, finite element analysis (FEM) was used to simulate structural adjustments at the converter station based on optimized protection parameters. For cable connection points, simulating seawater flow at a velocity of 0.8 m / s and a temperature of 22°C, the pit depth in stress concentration areas was reduced to 0.9 mm. After structural design adjustments, such as reinforcing weld seams at connection points, the overall structural stability improved. Finally, the failure probability distribution was updated, reducing the failure probability of cable connection points from 18% to 12%, and the failure probability of other components, such as support piles, from 10% to 8%. This optimized probability distribution data provides precise guidance for converter station maintenance. It should be noted that the extraction and parameter optimization of salinity fluctuation nonlinear characteristics, combined with FEM and probability updates, form a closed-loop analysis process. This method, supported by multi-dimensional data, ensures the accuracy of corrosion risk assessment and provides a reliable basis for the long-term operation of the converter station.

[0138] Example 2:

[0139] like Figure 2As shown in the figure, Embodiment 2 provides a structural optimization system for an offshore converter station based on simulation, including a simulation module 10, a dominant influencing factor determination module 20, a data analysis module 30, a distribution map generation module 40, and a structural optimization module 50.

[0140] The simulation module 10 is used to perform corrosion simulation on the target converter station through a virtual simulation platform based on the service environment data of the target converter station, and obtain a simulation dataset under the service environment.

[0141] The dominant influencing factor determination module 20 is used to analyze the interaction effect between temperature periodic gradient and stress corrosion coefficient based on the simulation dataset and a preset multi-field coupling analysis model, thereby determining the dominant influencing factor affecting the corrosion rate.

[0142] The data analysis module 30 is used to track the dynamic change characteristics of the pit depth threshold during the corrosion process using time series analysis methods based on the dominant influencing factor and the simulation dataset, and to obtain the distribution law data of the pit depth threshold under tidal cycle changes.

[0143] The distribution map generation module 40 is used to generate a failure probability distribution map of the target converter station based on the distribution pattern data.

[0144] The structural optimization module 50 is used to generate a dynamic protection parameter set based on the stress corrosion coefficient and fatigue corrosion factor according to the failure probability distribution map, and to perform structural optimization on the target converter station based on the dynamic protection parameter set.

[0145] Furthermore, the simulation module 10, based on the service environment data of the target converter station, performs corrosion simulation on the target converter station through a virtual simulation platform to obtain a corrosion characteristic dataset under the service environment, including:

[0146] The tidal cycle variation data, salinity fluctuation characteristic data, dissolved oxygen content distribution data, temperature cycle gradient and marine organism attachment data of the target converter station are acquired to construct the service environment data.

[0147] Based on the service environment data, a multi-field coupling model of the preset material under the service environment is constructed through a virtual simulation platform to obtain an initial environmental parameter set;

[0148] Using the finite element analysis method, based on the initial environmental parameter set, the stress distribution and electrochemical reaction of the preset material under a multi-field coupling environment are simulated to obtain a surface corrosion distribution map;

[0149] Based on the surface corrosion distribution map, the initial environmental parameter set is adjusted through an iterative optimization algorithm to obtain the optimized corrosion rate dataset and pit depth threshold dataset.

[0150] The corrosion characteristic dataset is constructed based on the corrosion rate dataset and the pit depth threshold dataset.

[0151] The service environment data and the corrosion characteristic dataset are spatiotemporally aligned to construct the corrosion characteristic dataset.

[0152] In one possible implementation, the dominant influencing factor determination module 20, based on the simulation dataset, analyzes the interaction effect between the temperature periodic gradient and the stress corrosion coefficient using a preset multi-field coupling analysis model, thereby determining the dominant influencing factor affecting the corrosion rate, including:

[0153] The temperature cycle gradient and stress corrosion coefficient of the service environment are determined based on the simulation dataset.

[0154] The interaction between the temperature periodic gradient and the stress corrosion coefficient is analyzed by the multi-field coupling analysis model to obtain the basic mapping relationship between different temperatures and different stress corrosion coefficients.

[0155] Based on the basic mapping relationship and the simulation dataset, the correlation mapping between each influencing factor and the corrosion rate is constructed respectively, and then the influence ranking of each influencing factor is determined. The influencing factors include wave impact frequency, biological attachment cycle, temperature cycle gradient, stress corrosion coefficient and salinity fluctuation amplitude.

[0156] Based on the ranking of the influence of each of the aforementioned influencing factors and the preset number of dominant influencing factors, several dominant influencing factors are determined.

[0157] In one possible implementation, the data analysis module 30, based on the dominant influencing factor and the simulation dataset, uses time series analysis to track the dynamic changes in the pit depth threshold during corrosion, obtaining data on the distribution pattern of the pit depth threshold under tidal cycle variations, including:

[0158] Extract the sub-datasets corresponding to each of the dominant influencing factors from the simulation dataset;

[0159] Based on the simulation dataset, the time series analysis method is used to preliminarily decompose the changes in the pit depth threshold and obtain the time series distribution of dynamic change characteristics.

[0160] Based on the tidal cycle variation data and the temporal distribution of the dynamic variation characteristics in the simulation dataset, an initial periodic force influence model is constructed.

[0161] Based on the initial periodic force influence model, the correlation features between periodic force and pit depth threshold change are extracted from the simulation dataset to obtain the initial distribution law data of pit depth threshold under tidal cycle change.

[0162] Based on each of the aforementioned subsets and the initial distribution pattern data, the support vector machine method is used to determine the influence coefficients of each of the aforementioned dominant influencing factors on the change of pit depth threshold.

[0163] The initial distribution data is corrected based on the respective influence coefficients and the respective subsets to obtain the distribution data of the pit depth threshold under tidal cycle changes.

[0164] In one possible implementation, the distribution map generation module 40 generates a failure probability distribution map of the target converter station based on the distribution pattern data, including:

[0165] Based on the distribution pattern data, the salinity fluctuation characteristic data and dissolved oxygen content distribution data in the simulation dataset, a mapping relationship model between the pit depth threshold and the marine corrosion cycle is constructed.

[0166] Based on the mapping relationship model, the simulation dataset, and the operating environment parameters of the target converter station, the degree of corrosion of each part of the target converter station is determined using the k-means clustering algorithm.

[0167] The corrosion rate dataset in the simulation dataset is weighted and updated according to the corrosion degree of each part of the target converter station to obtain the corrosion rate distribution data of each part of the target converter station.

[0168] By using the Monte Carlo simulation method, combined with the corrosion rate distribution data and each of the dominant influencing factors, failure probability distribution data for each part of the target converter station are generated.

[0169] The failure probability distribution data is plotted as a two-dimensional curve using visualization tools to obtain the failure probability distribution map.

[0170] Furthermore, the step of constructing a mapping model between the pit depth threshold and the marine corrosion cycle based on the distribution pattern data, the salinity fluctuation characteristic data and dissolved oxygen content distribution data in the simulation dataset includes:

[0171] The distribution data is decomposed into a time series to obtain a periodically changing sequence;

[0172] Based on the salinity fluctuation characteristic data, the dissolved oxygen content distribution data, and the periodic change sequence, an autoregressive moving average model is used for time series prediction to obtain the predicted periodic sequence. The autoregressive moving average model is a linear combination of autoregressive terms and moving average terms based on historical data, and is constructed after determining the model parameters through maximum likelihood estimation.

[0173] Based on the predicted cycle sequence and the simulation dataset, a mapping model between the pit depth threshold and the corrosion cycle is constructed by fitting the nonlinear relationship between the pit depth threshold and the corrosion cycle using a polynomial function.

[0174] In one possible implementation, the structural optimization module 50 generates a dynamic protection parameter set based on the stress corrosion coefficient and fatigue corrosion factor according to the failure probability distribution map, including:

[0175] Based on the failure probability distribution map, multiple rounds of numerical simulations were performed on the target converter station under the service environment to obtain the numerical simulation dataset and the nonlinear characteristic sequence of salinity fluctuations.

[0176] The preset medium corrosion index and preset electrochemical reaction parameters are adjusted according to the nonlinear characteristic sequence of the salinity fluctuation to obtain a set of corrosion parameters. The preset medium corrosion index and preset electrochemical reaction parameters are determined based on several corrosion experiments in a marine environment.

[0177] Estimate the numerical simulation dataset and the corrosion parameter set, and use regression analysis to construct a quantitative relationship expression between the pit depth threshold and the marine corrosion cycle, thereby determining the corrosion cycle model;

[0178] Based on the corrosion cycle model, the initial stress corrosion coefficient and the initial fatigue corrosion factor are optimized using a genetic algorithm to obtain the dynamic protection parameter set. The initial stress corrosion coefficient and the initial fatigue corrosion factor are determined based on the numerical simulation dataset. The dynamic protection parameter set includes material strength distribution data, material thickness distribution data, and connection stiffness distribution data.

[0179] Furthermore, the structure optimization module 50 performs structural optimization on the target converter station based on the dynamic protection parameter set, including:

[0180] Construct an initial structural parameter matrix based on the dynamic protection parameter set;

[0181] The initial structural parameter matrix is ​​iteratively updated several times using the finite element analysis method to obtain the final optimized structural parameter matrix. In each iteration, based on the current structural parameter matrix, a corresponding current structural finite element mesh model is generated using the finite element analysis method, and the current stress distribution and current deformation are calculated to obtain the current structural response data. Based on the current structural response data, the material strength distribution data and material thickness distribution data in the current structural parameter matrix are adjusted to obtain an updated structural parameter matrix, and the updated structural parameter matrix is ​​then iteratively updated again.

[0182] Based on the optimized structural parameter matrix, a structural adjustment scheme is generated, which includes the thickness adjustment value and connection stiffness adjustment value of specific components;

[0183] The target converter station is structurally optimized according to the structural adjustment scheme.

[0184] This application provides a simulation-based structural optimization system for offshore converter stations. First, a high-precision virtual simulation platform is used to perform a full-process, dynamic, and digital simulation of the corrosion behavior of the target converter station under its actual service environment, overcoming the limitations of traditional single, static anti-corrosion measures. Then, by introducing a multi-field coupling analysis model, the complex interaction effects between key physical fields such as temperature and stress are scientifically analyzed, accurately identifying the key influencing factors that truly dominate the corrosion process at different spatiotemporal scales, rather than relying on empirical judgment. Furthermore, time series analysis is applied to the dynamic tracking of pit depth thresholds, revealing the evolution law of corrosion damage under the action of periodic environmental forces such as tides, and generating an intuitive failure probability distribution map based on this. Ultimately, targeted and refined optimization of the converter station structure is achieved, constructing a complete technical closed loop from "virtual environment simulation" to "physical structure optimization," significantly improving the safety and reliability of offshore converter stations during long-term service in harsh marine environments.

[0185] For a more detailed explanation of the working principle and procedures of this embodiment, please refer to the relevant description in Embodiment 1.

[0186] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of this application. It should be understood that the above descriptions are merely specific embodiments of this application and are not intended to limit the scope of protection of this application. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application for those skilled in the art.

Claims

1. A structural optimization method for offshore converter stations based on simulation, characterized in that, include: Based on the service environment data of the target converter station, corrosion simulation is performed on the target converter station through a virtual simulation platform to obtain a simulation dataset under the service environment. Based on the simulation dataset, the interaction effect between temperature periodic gradient and stress corrosion coefficient is analyzed through a preset multi-field coupling analysis model, thereby determining the dominant influencing factor affecting the corrosion rate. Based on the dominant influencing factors and the simulation dataset, time series analysis was used to track the dynamic changes of the pit depth threshold during the corrosion process, and the distribution data of the pit depth threshold under tidal cycle changes were obtained. Based on the distribution pattern data, a failure probability distribution map of the target converter station is generated; Based on the failure probability distribution map, a dynamic protection parameter set based on the stress corrosion coefficient and fatigue corrosion factor is generated, and the target converter station is structurally optimized based on the dynamic protection parameter set.

2. The structural optimization method for a marine converter station based on simulation as described in claim 1, characterized in that, The step involves using a virtual simulation platform to perform corrosion simulation on the target converter station based on its service environment data, thereby obtaining a dataset of corrosion characteristics under the service environment, including: The tidal cycle variation data, salinity fluctuation characteristic data, dissolved oxygen content distribution data, temperature cycle gradient and marine organism attachment data of the target converter station are acquired to construct the service environment data. Based on the service environment data, a multi-field coupling model of the preset material under the service environment is constructed through a virtual simulation platform to obtain an initial environmental parameter set; Using the finite element analysis method, based on the initial environmental parameter set, the stress distribution and electrochemical reaction of the preset material under a multi-field coupling environment are simulated to obtain a surface corrosion distribution map; Based on the surface corrosion distribution map, the initial environmental parameter set is adjusted through an iterative optimization algorithm to obtain the optimized corrosion rate dataset and pit depth threshold dataset. The corrosion characteristic dataset is constructed based on the corrosion rate dataset and the pit depth threshold dataset. The service environment data and the corrosion characteristic dataset are spatiotemporally aligned to construct the corrosion characteristic dataset.

3. The structural optimization method for a marine converter station based on simulation as described in claim 1, characterized in that, Based on the simulation dataset, a pre-defined multi-field coupling analysis model is used to analyze the interaction effect between the temperature periodic gradient and the stress corrosion coefficient, thereby determining the dominant influencing factors affecting the corrosion rate, including: The temperature cycle gradient and stress corrosion coefficient of the service environment are determined based on the simulation dataset. The interaction between the temperature periodic gradient and the stress corrosion coefficient is analyzed by the multi-field coupling analysis model to obtain the basic mapping relationship between different temperatures and different stress corrosion coefficients. Based on the basic mapping relationship and the simulation dataset, the correlation mapping between each influencing factor and the corrosion rate is constructed respectively, and then the influence ranking of each influencing factor is determined. The influencing factors include wave impact frequency, biological attachment cycle, temperature cycle gradient, stress corrosion coefficient and salinity fluctuation amplitude. Based on the ranking of the influence of each of the aforementioned influencing factors and the preset number of dominant influencing factors, several dominant influencing factors are determined.

4. The structural optimization method for a marine converter station based on simulation as described in claim 1, characterized in that, The process involves using time series analysis based on the dominant influencing factors and the simulation dataset to track the dynamic changes in the pit depth threshold during corrosion, thereby obtaining data on the distribution pattern of the pit depth threshold under tidal cycle variations. This includes: Extract the sub-datasets corresponding to each of the dominant influencing factors from the simulation dataset; Based on the simulation dataset, the time series analysis method is used to preliminarily decompose the changes in the pit depth threshold and obtain the time series distribution of dynamic change characteristics. Based on the tidal cycle variation data and the temporal distribution of the dynamic variation characteristics in the simulation dataset, an initial periodic force influence model is constructed. Based on the initial periodic force influence model, the correlation features between periodic force and pit depth threshold change are extracted from the simulation dataset to obtain the initial distribution law data of pit depth threshold under tidal cycle change. Based on each of the aforementioned subsets and the initial distribution pattern data, the support vector machine method is used to determine the influence coefficients of each of the aforementioned dominant influencing factors on the change of pit depth threshold. The initial distribution data is corrected based on the respective influence coefficients and the respective subsets to obtain the distribution data of the pit depth threshold under tidal cycle changes.

5. The structural optimization method for a marine converter station based on simulation as described in claim 1, characterized in that, The step of generating a failure probability distribution map of the target converter station based on the distribution pattern data includes: Based on the distribution pattern data, the salinity fluctuation characteristic data and dissolved oxygen content distribution data in the simulation dataset, a mapping relationship model between the pit depth threshold and the marine corrosion cycle is constructed. Based on the mapping relationship model, the simulation dataset, and the operating environment parameters of the target converter station, the degree of corrosion of each part of the target converter station is determined using the k-means clustering algorithm. The corrosion rate dataset in the simulation dataset is weighted and updated according to the corrosion degree of each part of the target converter station to obtain the corrosion rate distribution data of each part of the target converter station. By using the Monte Carlo simulation method, combined with the corrosion rate distribution data and each of the dominant influencing factors, failure probability distribution data for each part of the target converter station are generated. The failure probability distribution data is plotted as a two-dimensional curve using visualization tools to obtain the failure probability distribution map.

6. The structural optimization method for an offshore converter station based on simulation as described in claim 5, characterized in that, The step of constructing a mapping model between the pit depth threshold and the marine corrosion cycle based on the distribution pattern data, the salinity fluctuation characteristic data and dissolved oxygen content distribution data in the simulation dataset includes: The distribution data is decomposed into a time series to obtain a periodically changing sequence; Based on the salinity fluctuation characteristic data, the dissolved oxygen content distribution data, and the periodic change sequence, an autoregressive moving average model is used for time series prediction to obtain the predicted periodic sequence. The autoregressive moving average model is a linear combination of autoregressive terms and moving average terms based on historical data, and is constructed after determining the model parameters through maximum likelihood estimation. Based on the predicted cycle sequence and the simulation dataset, a mapping model between the pit depth threshold and the corrosion cycle is constructed by fitting the nonlinear relationship between the pit depth threshold and the corrosion cycle using a polynomial function.

7. The structural optimization method for a marine converter station based on simulation as described in claim 1, characterized in that, The step of generating a dynamic protection parameter set based on the failure probability distribution map, including the following: Based on the failure probability distribution map, multiple rounds of numerical simulations were performed on the target converter station under the service environment to obtain the numerical simulation dataset and the nonlinear characteristic sequence of salinity fluctuations. The preset medium corrosion index and preset electrochemical reaction parameters are adjusted according to the nonlinear characteristic sequence of the salinity fluctuation to obtain a set of corrosion parameters. The preset medium corrosion index and preset electrochemical reaction parameters are determined based on several corrosion experiments in a marine environment. Estimate the numerical simulation dataset and the corrosion parameter set, and use regression analysis to construct a quantitative relationship expression between the pit depth threshold and the marine corrosion cycle, thereby determining the corrosion cycle model; Based on the corrosion cycle model, the initial stress corrosion coefficient and the initial fatigue corrosion factor are optimized using a genetic algorithm to obtain the dynamic protection parameter set. The initial stress corrosion coefficient and the initial fatigue corrosion factor are determined based on the numerical simulation dataset. The dynamic protection parameter set includes material strength distribution data, material thickness distribution data, and connection stiffness distribution data.

8. The structural optimization method for an offshore converter station based on simulation as described in claim 1, characterized in that, The structural optimization of the target converter station based on the dynamic protection parameter set includes: Construct an initial structural parameter matrix based on the dynamic protection parameter set; The initial structural parameter matrix is ​​iteratively updated several times using the finite element analysis method to obtain the final optimized structural parameter matrix. In each iteration, based on the current structural parameter matrix, a corresponding current structural finite element mesh model is generated using the finite element analysis method, and the current stress distribution and current deformation are calculated to obtain the current structural response data. Based on the current structural response data, the material strength distribution data and material thickness distribution data in the current structural parameter matrix are adjusted to obtain an updated structural parameter matrix, and the updated structural parameter matrix is ​​then iteratively updated again. Based on the optimized structural parameter matrix, a structural adjustment scheme is generated, which includes the thickness adjustment value and connection stiffness adjustment value of specific components; The target converter station is structurally optimized according to the structural adjustment scheme.

9. A structural optimization system for an offshore converter station based on simulation, characterized in that, It includes a simulation module, a module for determining dominant influencing factors, a data analysis module, a distribution map generation module, and a structure optimization module; The simulation module is used to perform corrosion simulation on the target converter station through a virtual simulation platform based on the service environment data of the target converter station, and obtain a simulation dataset under the service environment. The dominant influencing factor determination module is used to analyze the interaction effect between temperature periodic gradient and stress corrosion coefficient based on the simulation dataset and a preset multi-field coupling analysis model, thereby determining the dominant influencing factor affecting the corrosion rate. The data analysis module is used to track the dynamic change characteristics of the pit depth threshold during the corrosion process using time series analysis methods based on the dominant influencing factors and the simulation dataset, and to obtain the distribution pattern data of the pit depth threshold under tidal cycle changes. The distribution map generation module is used to generate a failure probability distribution map of the target converter station based on the distribution pattern data; The structural optimization module is used to generate a dynamic protection parameter set based on the stress corrosion coefficient and fatigue corrosion factor according to the failure probability distribution map, and to perform structural optimization on the target converter station based on the dynamic protection parameter set.

10. The structural optimization system for a marine converter station based on simulation as described in claim 9, characterized in that, The simulation module performs corrosion simulation on the target converter station using a virtual simulation platform based on the target converter station's service environment data, obtaining a dataset of corrosion characteristics under the service environment, including: The tidal cycle variation data, salinity fluctuation characteristic data, dissolved oxygen content distribution data, temperature cycle gradient and marine organism attachment data of the target converter station are acquired to construct the service environment data. Based on the service environment data, a multi-field coupling model of the preset material under the service environment is constructed through a virtual simulation platform to obtain an initial environmental parameter set; Using the finite element analysis method, based on the initial environmental parameter set, the stress distribution and electrochemical reaction of the preset material under a multi-field coupling environment are simulated to obtain a surface corrosion distribution map; Based on the surface corrosion distribution map, the initial environmental parameter set is adjusted through an iterative optimization algorithm to obtain the optimized corrosion rate dataset and pit depth threshold dataset. The corrosion characteristic dataset is constructed based on the corrosion rate dataset and the pit depth threshold dataset. The service environment data and the corrosion characteristic dataset are spatiotemporally aligned to construct the corrosion characteristic dataset.

Images