Method, device and equipment for relieving corrosion of metal structure of offshore converter station and medium

Abstract

The invention discloses a corrosion relieving method, device, equipment and medium for a metal structure of an offshore converter station, and the method comprises the steps: continuously collecting the real-time vibration data of a corresponding metal structure for a high-corrosion risk region of the metal structure of a target offshore converter station, and obtaining a vibration time sequence; inputting the vibration time sequence into a preset crack initiation prediction model to generate crack initiation distribution data, and predicting crack expansion data according to the crack initiation distribution data; and according to the crack expansion data, determining protective measure strength distribution, and according to the protective measure strength distribution and current environmental parameters, determining a metal structure corrosion mitigation scheme of the target offshore converter station. According to the scheme, accurate prediction of crack risks and dynamic optimization of protection parameters can be realized by fusing multi-physical field simulation and real-time data analysis, so that the reliability and durability of the metal structure of the offshore converter station are improved.

Description

Technical Field

[0001] This invention relates to the field of intelligent offshore converter station technology, and in particular to a method, apparatus, equipment and medium for mitigating corrosion of the metal structure of an offshore converter station. Background Technology

[0002] Offshore converter stations, as key facilities for offshore wind power transmission, are constantly exposed to the harsh marine environment of high salt spray and high humidity. Their metal structures not only face the threat of chemical corrosion but are also affected by stress concentration caused by mechanical and electromagnetic vibrations during operation. The coupled effect of these two factors significantly accelerates the initiation and propagation of fatigue cracks, seriously threatening the operational safety, structural lifespan, and wind power transmission efficiency of the converter station. Therefore, developing mitigation technologies that comprehensively consider the interaction between vibration and corrosion is of significant strategic importance for ensuring the long-term reliable operation of offshore energy infrastructure.

[0003] Current technologies for corrosion protection of metal structures in offshore converter stations often focus on single environmental factors or static protection methods, such as anti-corrosion coatings or cathodic protection, lacking a systematic analysis of the dynamic coupling mechanism between vibration characteristics and corrosion processes during equipment operation. Existing methods often fail to effectively identify the dynamic impact of vibration frequency tuning on structural stress distribution, nor do they perform time-series correlation analysis between real-time vibration data and corrosion rates. This results in an inability to accurately predict crack initiation and propagation trends, and protective measures lack specificity, foresight, and adaptability. Summary of the Invention

[0004] This invention provides a method, device, equipment, and medium for mitigating corrosion of metal structures in offshore converter stations. By integrating multiphysics simulation and real-time data analysis, it can achieve accurate prediction of crack risk and dynamic optimization of protection parameters, thereby improving the reliability and durability of metal structures in offshore converter stations.

[0005] In a first aspect, embodiments of the present invention provide a method for mitigating corrosion of the metal structure of an offshore converter station, comprising:

[0006] For high corrosion risk areas of the metal structure of the target offshore converter station, real-time vibration data of the corresponding metal structure is continuously collected to obtain a vibration time series. The high corrosion risk areas are determined by analyzing the erosion process of the metal structure in the stress concentration area of ​​the target offshore converter station under the influence of the marine environment. The stress concentration areas are determined by analyzing the vibration of the metal structure at different equipment tuning frequencies.

[0007] The vibration time series is input into a preset crack initiation prediction model to generate crack initiation distribution data, and crack propagation data is predicted based on the crack initiation distribution data; wherein, the crack initiation prediction model is constructed based on the temporal correlation between vibration changes and corrosion rate and Paris's law;

[0008] Based on the crack propagation data, the intensity distribution of protective measures is determined, and the current environmental parameters of the target offshore converter station are obtained. Based on the intensity distribution of protective measures and the current environmental parameters, a corrosion mitigation scheme for the metal structure of the target offshore converter station is determined.

[0009] This invention focuses on key risk areas, avoiding blind monitoring, reducing data redundancy, and ensuring the relevance and effectiveness of monitoring data, providing accurate input for subsequent crack prediction. By quantifying the coupling effect of vibration and corrosion through a model, it accurately obtains the initiation location, distribution, and propagation trend of cracks, addressing the pain point of being unable to predict crack development and providing data support for the formulation of protective measures. By precisely matching protective measures with crack risk and adapting to the real-time marine environment, it avoids insufficient or excessive protection, ensuring the practicality and effectiveness of mitigation solutions. Compared with existing technologies, this invention can achieve accurate prediction of crack risk and dynamic optimization of protection parameters by integrating multiphysics simulation and real-time data analysis, thereby improving the reliability and durability of the metal structure of offshore converter stations.

[0010] Furthermore, stress concentration regions are determined by analyzing the vibration of the metal structure at different device tuning frequencies, specifically:

[0011] By using a pre-established finite element model, the mechanical and electromagnetic vibrations of the metal structure under different device tuning frequencies are simulated to obtain vibration characteristic data and structural deformation data.

[0012] Using principal component analysis algorithm, the structural displacement field is calculated based on the vibration characteristic data and structural deformation data to generate a displacement distribution map;

[0013] The electromagnetic field data used in the finite element model simulation is superimposed on the displacement distribution map to obtain the composite stress field, and the stress concentration region is determined based on the composite stress field.

[0014] This invention recreates the actual vibration scenario of converter station equipment during operation using a finite element model, avoiding the bias of simulating a single vibration type and ensuring the comprehensiveness of vibration data. By reducing the dimensionality of complex vibration and deformation data, key features are retained, simplifying calculations while ensuring the accuracy of displacement field analysis. By superimposing electromagnetic field data, the influence of electromagnetic vibration on structural stress is considered, avoiding the stress analysis bias caused by focusing only on mechanical vibration, accurately locating stress concentration areas, and laying the foundation for subsequent identification of high corrosion risk areas.

[0015] Furthermore, by analyzing the erosion process of the metal structure in the stress concentration area of ​​the target offshore converter station under the influence of the marine environment, high corrosion risk areas were identified, specifically:

[0016] For the stress concentration area, a pre-constructed corrosion kinetic model is used to simulate the corrosion process of the metal structure under the influence of the marine environment, and an corrosion rate sequence is obtained. The corrosion rate characteristics are then extracted from the corrosion rate sequence. The corrosion rate sequence includes corrosion rate subsequences dominated by different marine environmental factors.

[0017] Based on the erosion rate characteristics and the preset three-dimensional geometric model of the corresponding erosion area, a depth distribution feature map of the erosion area is determined, and a composite damage field is obtained based on the depth distribution feature map of the erosion area and the composite stress field corresponding to the erosion area; wherein, the composite damage field is used to characterize the corrosion-stress synergistic damage mechanism.

[0018] Based on preset corrosion risk classification indicators, high corrosion risk areas are identified from the composite damage field; wherein, the corrosion risk classification indicators are determined by the K-nearest neighbor algorithm based on a preset marine environmental corrosion database.

[0019] This invention simulates the erosion process using a corrosion kinetic model, quantifies the erosion effects of marine environments such as salt spray and humidity on metals, distinguishes the influence weights of different environmental factors, and ensures the precision of erosion analysis. By combining a three-dimensional model of the eroded area and a composite stress field, it quantifies the synergistic damage mechanism of corrosion and stress, avoiding misjudgments of risks caused by analyzing corrosion or stress alone. The K-nearest neighbor algorithm is used to classify risk levels based on historical corrosion data, avoiding the bias of subjective threshold setting and accurately identifying high-risk areas that need to be monitored.

[0020] Furthermore, a crack initiation prediction model is constructed based on the temporal correlation between vibration changes and corrosion rates and Paris's law, specifically as follows:

[0021] The vibration amplitude change sequence and vibration frequency change sequence are obtained from a preset historical database, and the main frequency component and the corresponding amplitude data are extracted by Fourier transform based on the vibration amplitude change sequence and vibration frequency change sequence.

[0022] By using wavelet transform, the vibration amplitude variation data is determined based on the dominant frequency component and the corresponding amplitude data;

[0023] The corrosion rate time series at the same timestamp is obtained from the historical database, and the temporal correlation coefficient between the vibration amplitude change data and the corrosion rate time series is calculated by Pearson correlation analysis.

[0024] Based on the aforementioned temporal correlation coefficient and Paris's law, a crack initiation prediction model is constructed.

[0025] This invention employs Fourier transform to filter key frequency features from complex vibration time series, reducing irrelevant noise interference and improving the effectiveness of vibration data. By capturing the time-varying characteristics of vibration amplitude, it accurately reflects the dynamic changes in vibration, providing a foundation for establishing the correlation between vibration and corrosion. Pearson correlation is used to quantify the dynamic correlation strength between vibration changes and corrosion rates, clarifying the interaction between the two and providing core logic for model construction. Paris's law, a classic theory for crack initiation analysis, ensures that the model conforms to the physical laws of material crack initiation, improving prediction accuracy.

[0026] Furthermore, based on the crack initiation distribution data, crack propagation data is predicted, specifically as follows:

[0027] Crack size and location information are extracted from the crack initiation distribution data, and material property data and real-time stress field data of the corresponding metal structure are obtained based on the location information; wherein, the real-time stress field is obtained by vector superposition of electromagnetic stress field and vibration stress field;

[0028] Based on the material performance parameters and real-time stress field data, the stress intensity factor at the crack tip is calculated, and based on the stress intensity factor and a preset crack propagation rate model, the propagation path and rate of the crack in time and space are predicted, generating crack propagation data.

[0029] This invention extracts crack size and location information to obtain corresponding material properties and real-time stress field data, comprehensively collecting key influencing factors of crack propagation and avoiding prediction bias caused by single-factor analysis; it accurately quantifies the driving force of crack propagation by calculating the stress intensity factor; and it predicts the crack propagation path and rate by combining the crack propagation rate model, providing a precise basis for the formulation of protective measures.

[0030] Furthermore, based on the crack propagation data, the intensity distribution of protective measures is determined, specifically as follows:

[0031] The crack path distribution density is calculated based on the crack propagation data using the kernel density estimation method, thereby determining the crack distribution characteristics.

[0032] Using a pre-defined linear regression model, the strength values ​​of protective measures under different crack path distribution densities are calculated based on the crack distribution characteristics, thus obtaining the strength distribution of protective measures.

[0033] This invention uses a kernel density estimation method to determine crack distribution characteristics, visually presenting the concentrated areas and distribution patterns of cracks, and clarifying the key areas for protection. It also uses a linear regression model to calculate the strength values ​​of protective measures under different crack distribution densities, avoiding a mismatch between protection strength and risk (insufficient or excessive), ensuring the scientific nature of protection, and reducing protection costs.

[0034] Furthermore, based on the intensity distribution of the protective measures and the current environmental parameters, a corrosion mitigation scheme for the metal structure of the target offshore converter station is determined, specifically as follows:

[0035] Based on the intensity-thickness mapping table preset for the intensity distribution of the protective measures, the target protective coating thickness at the corresponding location of the metal structure is determined, and a spatial distribution scheme for the coating thickness is obtained.

[0036] The current environmental parameters are input into the pre-built protection current density calculation model to obtain the target average protection current density for the cathodic protection system. The target average protection current density is then corrected according to the intensity distribution of the protection measures to determine the final protection current density at the corresponding location of the metal structure, thus obtaining the final protection current density combination.

[0037] By integrating the spatial distribution scheme of coating thickness with the final protective current density, a corrosion mitigation scheme for metal structures is obtained.

[0038] This invention converts the protection strength into specific coating thickness parameters and adapts them to the risks of different areas according to spatial distribution, ensuring the targeted nature of coating protection. By calculating the target average protection current density based on the current environmental parameters and combining it with the protection strength distribution correction, the cathodic protection strength is adapted to the real-time environment and local crack risks, avoiding uneven protection caused by uniform current density.

[0039] Secondly, embodiments of the present invention provide a corrosion mitigation device for the metal structure of an offshore converter station, comprising a vibration data acquisition module, a crack data acquisition module, and a mitigation scheme acquisition module, wherein...

[0040] The vibration data acquisition module is used to continuously collect real-time vibration data of the corresponding metal structure in the high corrosion risk area of ​​the target offshore converter station, and obtain a vibration time sequence. The high corrosion risk area is determined by analyzing the erosion process of the metal structure in the stress concentration area of ​​the target offshore converter station under the influence of the marine environment. The stress concentration area is determined by analyzing the vibration of the metal structure at different equipment tuning frequencies.

[0041] The crack data acquisition module is used to input the vibration time series into a preset crack initiation prediction model to generate crack initiation distribution data, and predict crack propagation data based on the crack initiation distribution data; wherein, the crack initiation prediction model is constructed based on the temporal correlation between vibration changes and corrosion rates and Paris's law;

[0042] The mitigation scheme acquisition module is used to determine the intensity distribution of protective measures based on the crack propagation data, and to acquire the current environmental parameters of the target offshore converter station, so as to determine the corrosion mitigation scheme for the metal structure of the target offshore converter station based on the intensity distribution of protective measures and the current environmental parameters.

[0043] This invention employs a vibration data acquisition module to focus on key risk areas, avoiding blind monitoring, reducing data redundancy, and ensuring the relevance and effectiveness of monitoring data, providing accurate input for subsequent crack prediction. A crack data acquisition module quantifies the coupling effect of vibration and corrosion, accurately acquiring the initiation location, distribution, and propagation trend of cracks, addressing the pain point of being unable to predict crack development and providing data support for the formulation of protective measures. A mitigation solution acquisition module precisely matches protective measures with crack risks, while adapting to the real-time marine environment, avoiding insufficient or excessive protection, and ensuring the practicality and effectiveness of mitigation solutions.

[0044] Thirdly, embodiments of the present invention provide a terminal device, including: a processor, a memory, a communication interface, and a communication bus, wherein the processor, the memory, and the communication interface communicate with each other through the communication bus;

[0045] The memory is used to store at least one executable instruction that causes the processor to perform the operation of the corrosion mitigation method for the metal structure of the offshore converter station as described in any of the above.

[0046] Fourthly, embodiments of the present invention provide a computer-readable storage medium comprising a stored computer program, wherein, when the computer program is executed, it controls the device or apparatus containing the computer-readable storage medium to perform the corrosion mitigation method for the metal structure of an offshore converter station as described in any of the preceding claims.

[0047] The above description is merely an overview of the technical solutions of the embodiments of the present invention. In order to better understand the technical means of the embodiments of the present invention and to implement them in accordance with the contents of the specification, and to make the above and other objects, features and advantages of the embodiments of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0048] Figure 1This is a schematic diagram of a corrosion mitigation method for a metal structure of an offshore converter station provided in an embodiment of the present invention;

[0049] Figure 2 This is a structural diagram of a corrosion mitigation device for a metal structure of an offshore converter station, provided as an embodiment of the present invention. Detailed Implementation

[0050] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0051] Example 1:

[0052] like Figure 1 As shown, an embodiment of the present invention provides a method for mitigating corrosion of a metal structure in an offshore converter station, comprising the following steps:

[0053] S101, For the high corrosion risk area of ​​the metal structure of the target offshore converter station, real-time vibration data of the corresponding metal structure is continuously collected to obtain a vibration time sequence; wherein, the high corrosion risk area is determined by analyzing the erosion process of the metal structure in the stress concentration area of ​​the target offshore converter station under the influence of the marine environment; the stress concentration area is determined by analyzing the vibration of the metal structure at different equipment tuning frequencies;

[0054] In this embodiment, stress concentration regions are determined by analyzing the vibration of the metal structure at different device tuning frequencies. Specifically, a pre-established finite element model is used to simulate the mechanical and electromagnetic vibrations of the metal structure at different device tuning frequencies to obtain vibration characteristic data and structural deformation data. Principal component analysis is then used to calculate the structural displacement field based on the vibration characteristic data and structural deformation data to generate a displacement distribution map. The electromagnetic field data used in the finite element model simulation is superimposed onto the displacement distribution map to obtain a composite stress field, and the stress concentration regions are determined based on this composite stress field.

[0055] In one specific embodiment, a pre-established finite element model is used to input a tuning frequency to simulate mechanical and electromagnetic vibrations in the metal structure of an offshore converter station. Vibration characteristics are extracted from the simulation data to generate a vibration response sequence. Principal component analysis is used to process the vibration response sequence to obtain a reduced-dimensional vibration mode; the influence of vibration characteristics—vibration frequency—on the structure is analyzed through simulation. The structural displacement field is calculated based on the reduced-dimensional vibration mode, generating a displacement distribution map. If the deformation gradient in the displacement distribution map exceeds a preset threshold, it is determined that mechanical vibration has caused a local stress increase. The composite stress field is obtained by superimposing electromagnetic field data onto the displacement distribution map. The locations of stress concentration regions are identified from the composite stress field, and the intensity distribution map is determined.

[0056] Specifically, using a pre-established finite element model, the geometric model of the offshore converter station's metal structure is first constructed using ANSYS software, including the main beams, support columns, and connection nodes. The material properties are set as follows: steel elastic modulus of 2.1e11 Pa, Poisson's ratio of 0.3, and density of 7850 kg / m³. 3 The mesh is generated using tetrahedral elements, with a total of approximately 50,000 elements to ensure simulation accuracy.

[0057] Furthermore, mechanical vibrations at different tuning frequencies are simulated, for example, the tuning frequencies are set to 10Hz, 20Hz and 50Hz. Modal analysis algorithms are used to calculate the natural frequencies and modal shapes of the structure. Specifically, the process involves solving the generalized eigenvalue problem [K]{φ}=ω of the mass matrix [M] and stiffness matrix [K]. 2 [M]{φ}, where ω is the angular frequency and {φ} is the mode vector. The first 10 modes are solved iteratively by the Lanczos algorithm. The frequency of the first bending mode is 15.2Hz, which shows that the vibration amplitude is the smallest at 10Hz and close to resonance at 50Hz, resulting in a displacement amplification of 2.5 times.

[0058] Furthermore, electromagnetic vibration was simulated, and electromagnetic force was introduced as a load. Assuming an electromagnetic field strength of 0.5T and a force amplitude of 1000N, the Lorentz force F = J × B was calculated using the Maxwell equation coupled to the finite element method, where J is the current density 1e6A / m. 2 B is the magnetic induction intensity. The resonant excitation was simulated by a time stepping algorithm Δt = 0.001s, and the peak velocity of electromagnetic vibration at 20Hz was obtained as 0.05m / s.

[0059] Furthermore, the Fourier transform algorithm is used to perform frequency domain analysis on the time-domain vibration signal to extract the main frequency component and amplitude. For example, under 50Hz tuning, the main frequency peak is 48.7Hz and the amplitude is 0.12m. The data is stored in CSV format for easy subsequent processing.

[0060] Furthermore, the influence of vibration frequency on the structure was determined by comparing the stress response at different frequencies using Campbell's diagrams. The influence coefficient η = σ_max(ω) / σ_static was calculated, where σ_max is the maximum stress. At 10Hz, η = 1.2, while at 50Hz, η = 4.8, indicating that high-frequency vibration significantly amplifies the structural fatigue risk. Sensitivity analysis was then conducted to further investigate this effect. The quantization frequency sensitivity is 0.15 MPa / Hz;

[0061] Finally, the location and intensity distribution of the stress concentration area were obtained. Using the von Mises stress criterion, the root of the support column was identified as the concentration area, with a maximum stress of 250 MPa. The distribution cloud map showed that the intensity gradient decreased by 30% from the root to the top. The tuning frequency was adjusted to 25 Hz through optimization algorithms such as genetic algorithms to minimize the peak stress by 15%. This formed a complete logical chain from model construction to vibration simulation, data analysis, impact assessment, and area determination, ensuring that information technology automates the entire process.

[0062] In this embodiment, high corrosion risk areas are identified by analyzing the erosion process of the metal structure in the stress concentration area of ​​the target offshore converter station under the influence of the marine environment. Specifically, for the stress concentration area, the erosion process of the metal structure under the influence of the marine environment is simulated using a pre-constructed corrosion kinetic model to obtain an erosion rate sequence, and erosion rate features are extracted from the erosion rate sequence. The erosion rate sequence includes erosion rate sub-sequences dominated by different marine environmental factors. Based on the erosion rate features and a pre-defined three-dimensional geometric model of the corresponding erosion area, a depth distribution feature map of the erosion area is determined, and a composite damage field is obtained based on the depth distribution feature map of the erosion area and the composite stress field corresponding to the erosion area. The composite damage field is used to characterize the corrosion-stress synergistic damage mechanism. High corrosion risk areas are identified from the composite damage field according to a pre-defined corrosion risk classification index. The corrosion risk classification index is determined using the K-nearest neighbor algorithm based on a pre-defined marine environmental corrosion database.

[0063] In one specific embodiment, a pre-established corrosion kinetic model is used, inputting the location and intensity distribution data of stress concentration areas. Combined with salt spray and humidity factors, the corrosion process of metallic materials under environmental influences is simulated to obtain an initial corrosion rate sequence. If the fluctuation of the initial corrosion rate sequence exceeds a preset threshold, time series analysis is used to extract the periodic characteristics of the rate sequence, obtaining stable corrosion rate characteristics. Based on the stable corrosion rate characteristics, the finite difference method is used to calculate the corrosion depth distribution of the metallic material at different time steps, determining the corrosion depth data. Using the corrosion depth data and the geometric model of the metallic material, a three-dimensional distribution model of the corrosion region is constructed, obtaining the corrosion region distribution characteristics. If the maximum corrosion depth in the corrosion region distribution characteristics exceeds the material's allowable threshold, the Monte Carlo method is used to simulate the random changes of salt spray and humidity factors to determine the long-term stability of the corrosion rate. Based on the long-term stability judgment results, the parameters of the corrosion kinetic model are adjusted, and the corrosion process is re-simulated to obtain an optimized corrosion rate sequence. Using the optimized corrosion rate sequence and the intensity distribution of stress concentration areas, the remaining lifetime distribution of the metallic material is calculated, determining the final corrosion rate sequence. The salt spray-dominated and humidity-dominated erosion subsequences are extracted from the final erosion rate sequence. Principal component analysis (PCA) is used for dimensionality reduction to obtain a dimensionality-reduced erosion rate pattern. Based on the dimensionality-reduced erosion rate pattern, an interaction erosion map of salt spray and humidity factors is generated. If the erosion gradient in the interaction erosion map exceeds a preset threshold, the interaction area is marked as a preliminary high-risk zone. The stress location data is superimposed on the preliminary high-risk zone to obtain the corrosion stress composite field. The intensity distribution change of the corrosion stress composite field is judged. If the change exceeds a preset threshold, a composite high-risk zone is determined. Based on the composite high-risk zone, the data is input into a marine environmental corrosion database to retrieve historical corrosion data of similar metallic materials. The K-nearest neighbor algorithm is used for classification to obtain the classified corrosion risk level. Risk sub-levels related to the intensity distribution are extracted from the classified corrosion risk level to generate a risk distribution map and determine the coupling index sequence of corrosion factors and stress interaction. Based on the coupling index sequence, the weight coefficients of salt spray and humidity factors are calculated to obtain the final coupling index. If the final coupling index exceeds a preset threshold, a list of marked high-risk zones is output.

[0064] Specifically, based on previously identified data on the location and intensity distribution of stress concentration areas such as the base of the support columns, data was imported into a marine environmental parameter database, including a salt spray concentration set at 5.2 g / m³. 3With relative humidity maintained at 85%, the erosion rate k = Aexp(-Ea / RT) was first described using the Arrhenius equation in a corrosion kinetic model. The pre-exponential factor A was 1.8e8 mm / year, the activation energy Ea was 50 kJ / mol, the gas constant R was 8.314 J / mol·K, and the temperature T was 298 K. The base erosion rate under salt spray was calculated to be 0.12 mm / year. Subsequently, a humidity influence factor β = 1 + 0.02(humidity - 60%) was coupled in to obtain a corrected rate of 0.18 mm / year. The erosion depth evolution equation was then iteratively solved using the finite difference method on a grid with a time step Δt = 1 day. Where σ represents the local stress of 250 MPa and the yield strength σ_yield is 350 MPa, the root depth increment after 30 days of simulation reaches 0.015 mm. This rate is then compared with the preset threshold of 0.10 mm / year. If it exceeds, it is automatically marked as a high-risk area, for example, the risk level of the root area is raised to level 4 and the connection node to level 3. Uncertainty is assessed through Monte Carlo simulation. The risk probability distribution average is 72% after sampling 1000 salt spray fluctuations ±10%, providing a basis for subsequent protection design. On this basis, corrosion factors and stress interaction indicators are initially coupled, and a coupling index is defined. γ = ε_corrosion × (σ_max / σ_threshold), where ε_corrosion is the normalized corrosion rate of 0.65 and σ_threshold is 150 MPa. The calculated root γ = 1.08 exceeds the threshold of 1.0, indicating that the interaction effect amplifies the fatigue crack propagation rate by 2.1 times. Principal component analysis was used to extract the main interaction patterns from the multivariate dataset, with a variance contribution rate of 85%, thus quantifying the corrosion-stress synergistic damage mechanism. A neural network model was trained to predict the long-term coupled evolution. 100 sets of historical data were input, and the output R... 2 With a fitting accuracy of 0.92, an automated analysis chain is formed, from importing regional data to rate calculation, risk labeling, and determination of coupling indicators, ensuring that information technology processes the environmental interaction effects throughout the entire process.

[0065] S102, the vibration time sequence is input into a preset crack initiation prediction model to generate crack initiation distribution data, and crack propagation data is predicted based on the crack initiation distribution data; wherein, the crack initiation prediction model is constructed based on the temporal correlation between vibration change and corrosion rate and Paris's law;

[0066] In this embodiment, a crack initiation prediction model is constructed based on the temporal correlation between vibration changes and corrosion rates and Paris's law. Specifically, the following steps are taken: Vibration amplitude change sequences and vibration frequency change sequences are obtained from a preset historical database. Fourier transform is used to extract the dominant frequency component and corresponding amplitude data based on these sequences. Wavelet transform is used to determine the vibration amplitude change data based on the dominant frequency component and corresponding amplitude data. A corrosion rate time series at the same timestamp is obtained from the historical database. Pearson correlation analysis is used to calculate the temporal correlation coefficient between the vibration amplitude change data and the corrosion rate time series. Based on the temporal correlation coefficient and Paris's law, a crack initiation prediction model is constructed.

[0067] In one specific embodiment, for the preliminary coupling index, real-time equipment operation data, including vibration amplitude changes and frequency changes, are acquired. Vibration amplitude change sequences and frequency change sequences are extracted from the real-time equipment operation data. Fourier transform is used to process the vibration amplitude change sequences and frequency change sequences to obtain a frequency domain vibration spectrum. Based on the vibration spectrum, the dominant frequency component and its corresponding amplitude are extracted, and wavelet transform is used to analyze the characteristics of the dominant frequency component changing over time to determine the vibration amplitude change trend. The corrosion rate time series is obtained through electrochemical measurement. If the correlation coefficient between the corrosion rate and the vibration amplitude change time series exceeds a threshold within a preset time window, a temporal correlation is determined, and a preliminary correlation index is obtained. Pearson correlation analysis is used to calculate the temporal correlation coefficient between vibration amplitude change and corrosion rate to determine the quantitative correlation strength between the two. Based on the temporal correlation coefficient, a vibration-corrosion coupling model is constructed, where the coupling model refers to fitting the relationship between vibration amplitude change and corrosion rate using a linear regression method to obtain model parameters. The influence of corrosion rate on vibration amplitude is predicted using the coupling model, and a sliding time window is used to analyze the deviation between the predicted results and the actual data to determine the applicability of the model. Based on the model's applicability, the coupling model parameters are optimized. If the prediction deviation exceeds a preset threshold, the model parameters are adjusted to obtain the temporal correlation between vibration amplitude variation and corrosion rate. The correlation matrix between vibration amplitude and corrosion rate is then obtained based on this temporal correlation. Finally, a fatigue crack initiation model is constructed using the correlation matrix and Paris's model.

[0068] Specifically, for the preliminary coupling index γ = 1.08, real-time equipment operation data was acquired via an IoT sensor network, resulting in a time series with a peak vibration amplitude of 3.2 mm and a frequency variation range of 15-25 Hz. The data was then imported into a time series analysis algorithm using the ARIMA model for stationarization. First, the stationarity of the sequence was confirmed by an ADF test with a p-value of 0.023. Then, the parameters p = 2, d = 1, and q = 1 were determined using the autocorrelation function ACF and the partial autocorrelation function PACF. The fitting model coefficient φ1 = 0. .45, φ2=-0.32, θ1=0.28, the Ljung-Box statistic for residual white noise test Q=12.6p=0.19 indicates that the model is effective. Furthermore, the time-varying relationship of corrosion factors is calculated by introducing the lag term corrosion rate k(t-1)=0.16mm / year as an exogenous variable X_t, constructing a VAR vector autoregressive model with a lag order L=3, selected by the minimum value of the AIC criterion as -2.15. The coefficient matrix shows that the impulse response coefficient of vibration amplitude to corrosion rate is 0.021mm / year per The frequency change response coefficient to corrosion rate was -0.008 mm / year per Hz. The fatigue crack initiation acceleration mechanism under simulated interaction was investigated using the modified form of Paris's law: da / dN = C(ΔK)^m × exp(αk), where the crack propagation rate constant C = 1.5e-8 m / cycle, the Paris exponent m = 3.2, the stress intensity factor range ΔK = 18 MPa√m, and the corrosion acceleration factor α = 0.12. This was obtained by normalizing the coupling index γ. The Granger causality test (F statistic = 4.67, p = 0.001) confirmed that the vibration data significantly led to changes in the crack initiation rate due to Granger causality.

[0069] In this embodiment, crack propagation data is predicted based on the crack initiation distribution data. Specifically, crack size and location information are extracted from the crack initiation distribution data, and material property data and real-time stress field data of the corresponding metal structure are obtained based on the location information. The real-time stress field is obtained by vector superposition of electromagnetic stress field and vibration stress field. Based on the material property parameters and real-time stress field data, the stress intensity factor at the crack tip is calculated, and based on the stress intensity factor and a preset crack propagation rate model, the propagation path and rate of the crack in time and space are predicted to generate crack propagation data.

[0070] In one specific embodiment, crack features are read from the crack initiation distribution map using a data parsing algorithm (such as image processing or mesh scanning). These features include the initial crack length a0 (e.g., 0.1 mm), crack width (e.g., 0.05 mm), and location coordinates (e.g., the coordinates of the mesh node N101 at the base of the support column: (x, y, z) = (10.2, 5.6, 0.0) m). The extraction process employs a threshold segmentation method. If the pixel intensity of the crack region exceeds a preset threshold (e.g., 200 MPa equivalent stress), it is marked as a valid crack, and a crack list in CSV format is generated and stored, containing crack ID, size, and location fields.

[0071] Furthermore, material parameters for location N101 were retrieved from a pre-built material property database, including yield strength σ_yield = 350 MPa, fracture toughness K_IC = 50 MPa√m, elastic modulus E = 2.1e11 Pa, Poisson's ratio ν = 0.3, and density ρ = 7850 kg / m³. 3 Real-time stress field data is collected via an IoT sensor network, including vibration stress field σ_vib and electromagnetic stress field σ_em, at a sampling frequency of 100Hz in time-series format. A vector superposition algorithm is used to calculate the real-time stress field σ_total = σ_vib + σ_em, where σ_vib is converted based on accelerometer data (e.g., peak value 120MPa) and σ_em is converted based on electromagnetic force sensor data (e.g., peak value 80MPa). After superposition, σ_total = 200MPa. The data is transmitted in real-time to an edge computing device via a 5G network and stored in HDF5 format.

[0072] Furthermore, a local finite element model of the crack tip was constructed, and a refined mesh (0.01 mm element size, approximately 10,000 elements) was created using ANSYS software, with material properties set according to S502. The stress intensity factor range ΔK was calculated using the J-integral method or the direct displacement method. For surface cracks, the formula ΔK = Yσ_total√(πa) was used, where Y is the geometric correction factor (1.12 for semi-elliptical cracks), and a is the current crack length (0.1 mm). The calculated ΔK = 1.12 × 200 × √(3.14 × 0.0001) ≈ 12.5 MPa√m was obtained. The accuracy of ΔK was verified through parametric analysis; for example, by varying the crack angle from 0 to 90°, ΔK fluctuated within a range of 10-15 MPa√m.

[0073] Furthermore, a modified form of Paris's law is adopted as the crack propagation rate model: da / dN=C(ΔK)^m×exp(αk), where C is the crack propagation rate constant (1.5e-8m / cycle), m is the Paris exponent (3.2), α is the corrosion acceleration factor (0.1), and k is the corrosion rate (0.12mm / year). The calculated da / dN=1.5e-8×(12.5)^3.2×exp(0.1×0.12)≈1.5e-8×500×1.012≈7.59e-6m / cycle. Converting the number of cycles N to time t, based on the equipment operating frequency f=25Hz, then da / dt=da / dN×f=7.59e-6×25=1.8975e-4mm / s. The crack propagation path is predicted using numerical integration methods (such as the Euler method): a(t)=a0+∫0 t The crack length time series is generated by iterating over 30 days (720 hours) using the formula da / dt dt, with a time step Δt = 1 hour. For example, on day 30, a = 0.125 mm. Simultaneously, the crack propagation direction is simulated using finite element software, and a spatial path diagram is generated based on the maximum principal stress criterion (e.g., the crack propagates along a 45° direction, and the path coordinate sequence is stored in JSON format).

[0074] Furthermore, crack propagation data includes crack length time series, propagation path coordinates, rate curves, and risk indicators (such as failure time). Uncertainty was assessed using Monte Carlo simulation with 1000 samples and parameter fluctuation ranges of ΔK±5%, C±10%, and k±8%, yielding a probability distribution of the crack propagation data (e.g., a mean crack length of 0.125 mm, standard deviation of 0.005 mm, and a 95% confidence interval of [0.115, 0.135] mm after 30 days). The data was output to a cloud database for subsequent optimization of protective measures.

[0075] S103, Based on the crack propagation data, determine the intensity distribution of protective measures, obtain the current environmental parameters of the target offshore converter station, and determine the corrosion mitigation scheme for the metal structure of the target offshore converter station based on the intensity distribution of protective measures and the current environmental parameters.

[0076] In this embodiment, the intensity distribution of protective measures is determined based on the crack propagation data. Specifically, the crack path distribution density is calculated based on the crack propagation data using the kernel density estimation method to determine the crack distribution characteristics. The intensity values ​​of protective measures under different crack path distribution densities are calculated based on the crack distribution characteristics using a preset linear regression model to obtain the intensity distribution of protective measures.

[0077] In one specific embodiment, the spatial coordinate sequence of the crack path and the propagation rate data are extracted from the crack propagation data to construct a crack path dataset. Subsequently, the intensity distribution of protective measures is determined through the following two core steps:

[0078] S601. Extract the spatial coordinates (x, y, z) and propagation rate of the crack from the crack propagation data to form a dataset containing N data points. Use a Gaussian kernel function, optimize according to the Silverman rule, set the bandwidth parameter h = 0.1 mm, divide the metal structure surface into a 100×100 element mesh, and set the mesh resolution Δx = 0.5 mm. At each mesh point, apply the following kernel density estimation formula: f(x) = (1 / (N*h))*Σ[i=1 to N]K((x-X_i) / h), where f(x) is the crack path distribution density at mesh point x, N is the total number of crack path data points, K(·) is the Gaussian kernel function, X_i is the coordinate point of the i-th crack path, and h is the bandwidth parameter.

[0079] Furthermore, after the calculation is completed, a crack path distribution density map is generated. Density values ​​exceeding a preset threshold (e.g., ρ_threshold = 0.05 cracks / mm) are then considered. 2 The regions are marked as high-density areas to determine crack distribution characteristics, including density peaks, gradient changes, and spatial clustering patterns.

[0080] Furthermore, the protective measure intensity value S is calculated using the following multiple linear regression model, and the linear regression formula is:

[0081]

[0082] Where S is the calculated strength value of the protective measure (unit: MPa, which can characterize the equivalent compressive strength of the protective layer or the mapping value corresponding to the cathodic protection current density), and ρ is the crack path distribution density. β_0, β_1, and β_2 are the spatial gradient values ​​of the density, and the regression coefficients obtained by training with historical data (e.g., β_0 = 10.2, β_1 = 15.6, β_2 = -3.4).

[0083] Crack distribution characteristics (such as density ρ and gradient at each grid point) Input the above regression model and calculate the strength value S of the protective measures point by point.

[0084] Furthermore, after traversing all grid points, a strength distribution map of the protective measures for the entire metal structure is generated. If the strength value of a certain area is lower than the preset minimum strength threshold (e.g., S_min = 12MPa), the model parameters are automatically adjusted or interactive terms are introduced to optimize the model. The final output is strength distribution data that can be used to guide the formulation of protection schemes.

[0085] In this embodiment, based on the intensity distribution of the protective measures and the current environmental parameters, a corrosion mitigation scheme for the metal structure of the target offshore converter station is determined. Specifically, the target protective coating thickness at the corresponding location of the metal structure is determined according to a preset intensity-thickness mapping table based on the intensity distribution of the protective measures, resulting in a spatial distribution scheme for the coating thickness. The current environmental parameters are input into a pre-constructed protective current density calculation model to obtain the target average protective current density for the cathodic protection system. Based on the intensity distribution of the protective measures, the target average protective current density is corrected to determine the final protective current density at the corresponding location of the metal structure, resulting in a final protective current density combination. The spatial distribution scheme for the coating thickness is integrated with the final protective current density to obtain a corrosion mitigation scheme for the metal structure.

[0086] In one specific embodiment, based on the calculated distribution of protective measures intensity (intensity value S ranges from 0 to 1, where 0 represents no risk and 1 represents the highest risk), a pre-fitted intensity-thickness mapping table based on experimental data is invoked. This mapping relationship is defined by a quadratic polynomial: coating thickness T = 0.5 * S^2 + 0.3 * S + 0.1, in millimeters. For example, when the protective measures intensity value S = 0.8 in a certain area, the target coating thickness at that location is calculated as T = 0.5 * (0.8 squared) + 0.3 * 0.8 + 0.1 = 0.5 * 0.64 + 0.24 + 0.1 = 0.66 mm.

[0087] Furthermore, for the entire metal structure, based on the mesh nodes of the finite element model (resolution 0.1 meters), the corresponding coating thickness is calculated point by point, thereby generating a spatial distribution scheme of coating thickness covering the entire structure. This scheme is output in the form of data tables (such as CSV format) and 3D cloud maps, clearly indicating the coating thickness requirements for different areas. For example, the thickness in high-risk areas such as the base of the support columns is 0.66 mm, while the thickness in low-risk areas such as the top platform can be 0.1 mm.

[0088] Furthermore, a protective current density calculation model was established based on electrochemical kinetics principles. The model input consisted of current environmental parameters, including seawater salinity (35 g / L), temperature (15°C), pH (8.1), and dissolved oxygen concentration (8 mg / L). The model calculation yielded a target average protective current density I_target = 0.12 mA / cm² under standard conditions. 2 .

[0089] Furthermore, this average value is locally corrected based on the distribution of protective measure intensity. The correction formula is: Final protective current density I_final = I_target * (1 + 0.6 * S). Where S is the protective measure intensity value at a local location, and 0.6 is an empirical intensity correction coefficient. Taking the aforementioned region where S = 0.8 as an example, its final protective current density I_final = 0.12 * (1 + 0.6 * 0.8) = 0.12 * 1.48 = 0.1776 mA / cm² 2 Perform this calculation on all grid points to generate a final combination of protective current densities corresponding to the structural geometry, explicitly guiding the current output of the cathodic protection system at different locations.

[0090] Furthermore, the integration process is achieved through data fusion, linking the coating thickness and protective current density schemes obtained in the previous two steps to each grid cell of the metal structure, forming a complete, spatially differentiated protection strategy. For example, for the critical location at the base of the support column, the scheme explicitly specifies: protective coating thickness: 0.66 mm, cathodic protection current density: 0.178 mA / cm². 2 .

[0091] The final corrosion mitigation plan for the metal structure is a comprehensive document containing the following: coating construction drawings, indicating the specific thickness to be coated in each area; a cathodic protection configuration table, specifying the required protection current density for each area; an implementation priority list, determining the priority order of construction and maintenance based on risk level (e.g., areas with high strength value S>0.7); and a long-term monitoring and maintenance plan, for example, stipulating that the coating thickness be sampled and measured every 5 years and that the output current of the cathodic protection system be calibrated annually.

[0092] This invention focuses on key risk areas, avoiding blind monitoring, reducing data redundancy, and ensuring the relevance and effectiveness of monitoring data, providing accurate input for subsequent crack prediction. By quantifying the coupling effect of vibration and corrosion through a model, it accurately obtains the initiation location, distribution, and propagation trend of cracks, addressing the pain point of being unable to predict crack development and providing data support for the formulation of protective measures. By precisely matching protective measures with crack risk and adapting to the real-time marine environment, it avoids insufficient or excessive protection, ensuring the practicality and effectiveness of mitigation solutions. Compared with existing technologies, this invention can achieve accurate prediction of crack risk and dynamic optimization of protection parameters by integrating multiphysics simulation and real-time data analysis, thereby improving the reliability and durability of the metal structure of offshore converter stations.

[0093] Example 2:

[0094] like Figure 2As shown, this embodiment provides a corrosion mitigation device for the metal structure of an offshore converter station, including a vibration data acquisition module 201, a crack data acquisition module 202, and a mitigation scheme acquisition module 203.

[0095] The vibration data acquisition module 201 is used to continuously collect real-time vibration data of the corresponding metal structure in the high corrosion risk area of ​​the target offshore converter station's metal structure to obtain a vibration time sequence; wherein, the high corrosion risk area is determined by analyzing the erosion process of the metal structure in the stress concentration area of ​​the target offshore converter station under the influence of the marine environment; the stress concentration area is determined by analyzing the vibration of the metal structure at different equipment tuning frequencies;

[0096] The crack data acquisition module 202 is used to input the vibration time sequence into a preset crack initiation prediction model to generate crack initiation distribution data, and predict crack propagation data based on the crack initiation distribution data; wherein, the crack initiation prediction model is constructed based on the temporal correlation between vibration change and corrosion rate and Paris's law;

[0097] In this embodiment, the crack data acquisition module 202 predicts crack propagation data based on the crack initiation distribution data. Specifically, the crack data acquisition module 202 extracts crack size and location information from the crack initiation distribution data, and acquires material property data and real-time stress field data of the metal structure at the corresponding location based on the location information. The real-time stress field is obtained by vector superposition of electromagnetic stress field and vibration stress field. Based on the material property parameters and real-time stress field data, the stress intensity factor at the crack tip is calculated, and based on the stress intensity factor and a preset crack propagation rate model, the propagation path and rate of the crack in time and space are predicted to generate crack propagation data.

[0098] The mitigation scheme acquisition module 203 is used to determine the intensity distribution of protective measures based on the crack propagation data, acquire the current environmental parameters of the target offshore converter station, and determine the corrosion mitigation scheme for the metal structure of the target offshore converter station based on the intensity distribution of protective measures and the current environmental parameters.

[0099] In this embodiment, the mitigation scheme acquisition module 203 determines the intensity distribution of protective measures based on the crack propagation data. Specifically, the mitigation scheme acquisition module 203 calculates the crack path distribution density based on the crack propagation data using a kernel density estimation method to determine the crack distribution characteristics. Then, using a preset linear regression model, it calculates the intensity values ​​of protective measures under different crack path distribution densities based on the crack distribution characteristics to obtain the intensity distribution of protective measures.

[0100] In this embodiment, the mitigation scheme acquisition module 203 determines the corrosion mitigation scheme for the metal structure of the target offshore converter station based on the intensity distribution of the protective measures and the current environmental parameters. Specifically, the mitigation scheme acquisition module 203 determines the target protective coating thickness at the corresponding location of the metal structure based on a preset intensity-thickness mapping table of the intensity distribution of the protective measures, thus obtaining a spatial distribution scheme for the coating thickness; the current environmental parameters are input into a pre-constructed protective current density calculation model to obtain the target average protective current density for the cathodic protection system, and the target average protective current density is corrected based on the intensity distribution of the protective measures to determine the final protective current density at the corresponding location of the metal structure, thus obtaining a final protective current density combination; the spatial distribution scheme for the coating thickness is integrated with the final protective current density to obtain the corrosion mitigation scheme for the metal structure.

[0101] For a more detailed explanation of the working principle and procedures of this embodiment, please refer to the relevant description in Embodiment 1.

[0102] This invention employs a vibration data acquisition module 201 to focus on key risk areas, avoiding blind monitoring, reducing data redundancy, and ensuring the relevance and effectiveness of monitoring data, providing accurate input for subsequent crack prediction. A crack data acquisition module 202 quantifies the coupling effect of vibration and corrosion, accurately acquiring the initiation location, distribution, and propagation trend of cracks, addressing the pain point of being unable to predict crack development and providing data support for the formulation of protective measures. A mitigation solution acquisition module 203 precisely matches protective measures with crack risks, while adapting to the real-time marine environment, avoiding insufficient or excessive protection, and ensuring the practicality and effectiveness of mitigation solutions.

[0103] Example 3:

[0104] This embodiment provides a terminal device, including: a processor, a memory, a communication interface, and a communication bus, wherein the processor, the memory, and the communication interface communicate with each other through the communication bus;

[0105] The memory is used to store at least one executable instruction that causes the processor to perform the operation of the corrosion mitigation method for the metal structure of the offshore converter station as described in any of the above.

[0106] Example 4:

[0107] This invention provides a computer-readable storage medium comprising a stored computer program, wherein, when the computer program is executed, it controls the device or apparatus containing the computer-readable storage medium to perform the corrosion mitigation method for the metal structure of an offshore converter station as described in any of the preceding embodiments.

[0108] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0109] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. A method for mitigating corrosion of metal structures in offshore converter stations, characterized in that, include: For high corrosion risk areas of the metal structure of the target offshore converter station, real-time vibration data of the corresponding metal structure is continuously collected to obtain a vibration time series. The high corrosion risk areas are determined by analyzing the erosion process of the metal structure in the stress concentration area of ​​the target offshore converter station under the influence of the marine environment. The stress concentration areas are determined by analyzing the vibration of the metal structure at different equipment tuning frequencies. The vibration time series is input into a preset crack initiation prediction model to generate crack initiation distribution data, and crack propagation data is predicted based on the crack initiation distribution data; wherein, the crack initiation prediction model is constructed based on the temporal correlation between vibration changes and corrosion rate and Paris's law; Based on the crack propagation data, the intensity distribution of protective measures is determined, and the current environmental parameters of the target offshore converter station are obtained. Based on the intensity distribution of protective measures and the current environmental parameters, a corrosion mitigation scheme for the metal structure of the target offshore converter station is determined.

2. The corrosion mitigation method for the metal structure of an offshore converter station as described in claim 1, characterized in that, The stress concentration regions were determined by analyzing the vibration of the metal structure at different device tuning frequencies. Specifically: By using a pre-established finite element model, the mechanical and electromagnetic vibrations of the metal structure under different device tuning frequencies are simulated to obtain vibration characteristic data and structural deformation data. Using principal component analysis algorithm, the structural displacement field is calculated based on the vibration characteristic data and structural deformation data to generate a displacement distribution map; The electromagnetic field data used in the finite element model simulation is superimposed on the displacement distribution map to obtain the composite stress field, and the stress concentration region is determined based on the composite stress field.

3. The corrosion mitigation method for the metal structure of an offshore converter station as described in claim 2, characterized in that, By analyzing the erosion process of the metal structure in the stress concentration area of ​​the target offshore converter station under the influence of the marine environment, high corrosion risk areas were identified, specifically: For the stress concentration area, a pre-constructed corrosion kinetic model is used to simulate the corrosion process of the metal structure under the influence of the marine environment, and an corrosion rate sequence is obtained. The corrosion rate characteristics are then extracted from the corrosion rate sequence. The corrosion rate sequence includes corrosion rate subsequences dominated by different marine environmental factors. Based on the erosion rate characteristics and the preset three-dimensional geometric model of the corresponding erosion area, a depth distribution feature map of the erosion area is determined, and a composite damage field is obtained based on the depth distribution feature map of the erosion area and the composite stress field corresponding to the erosion area; wherein, the composite damage field is used to characterize the corrosion-stress synergistic damage mechanism. Based on preset corrosion risk classification indicators, high corrosion risk areas are identified from the composite damage field; wherein, the corrosion risk classification indicators are determined by the K-nearest neighbor algorithm based on a preset marine environmental corrosion database.

4. The corrosion mitigation method for the metal structure of an offshore converter station as described in claim 1, characterized in that, A crack initiation prediction model is constructed based on the temporal correlation between vibration changes and corrosion rates and Paris's law, specifically as follows: The vibration amplitude change sequence and vibration frequency change sequence are obtained from a preset historical database, and the main frequency component and the corresponding amplitude data are extracted by Fourier transform based on the vibration amplitude change sequence and vibration frequency change sequence. By using wavelet transform, the vibration amplitude variation data is determined based on the dominant frequency component and the corresponding amplitude data; The corrosion rate time series at the same timestamp is obtained from the historical database, and the temporal correlation coefficient between the vibration amplitude change data and the corrosion rate time series is calculated by Pearson correlation analysis. Based on the aforementioned temporal correlation coefficient and Paris's law, a crack initiation prediction model is constructed.

5. A method for mitigating corrosion of a marine converter station's metal structure as described in claim 1, characterized in that, Based on the crack initiation distribution data, the crack propagation data is predicted as follows: Crack size and location information are extracted from the crack initiation distribution data, and material property data and real-time stress field data of the corresponding metal structure are obtained based on the location information; wherein, the real-time stress field is obtained by vector superposition of electromagnetic stress field and vibration stress field; Based on the material performance parameters and real-time stress field data, the stress intensity factor at the crack tip is calculated, and based on the stress intensity factor and a preset crack propagation rate model, the propagation path and rate of the crack in time and space are predicted, generating crack propagation data.

6. The corrosion mitigation method for the metal structure of an offshore converter station as described in claim 1, characterized in that, Based on the crack propagation data, the intensity distribution of protective measures is determined as follows: The crack path distribution density is calculated based on the crack propagation data using the kernel density estimation method, thereby determining the crack distribution characteristics. Using a pre-defined linear regression model, the strength values ​​of protective measures under different crack path distribution densities are calculated based on the crack distribution characteristics, thus obtaining the strength distribution of protective measures.

7. The corrosion mitigation method for the metal structure of an offshore converter station as described in claim 1, characterized in that, Based on the intensity distribution of the protective measures and the current environmental parameters, a corrosion mitigation scheme for the metal structure of the target offshore converter station is determined, specifically as follows: Based on the intensity-thickness mapping table preset for the intensity distribution of the protective measures, the target protective coating thickness at the corresponding location of the metal structure is determined, and a spatial distribution scheme for the coating thickness is obtained. The current environmental parameters are input into the pre-built protection current density calculation model to obtain the target average protection current density for the cathodic protection system. The target average protection current density is then corrected according to the intensity distribution of the protection measures to determine the final protection current density at the corresponding location of the metal structure, thus obtaining the final protection current density combination. By integrating the spatial distribution scheme of coating thickness with the final protective current density, a corrosion mitigation scheme for metal structures is obtained.

8. A corrosion mitigation device for the metal structure of an offshore converter station, characterized in that, It includes a vibration data acquisition module, a crack data acquisition module, and a mitigation scheme acquisition module, among which, The vibration data acquisition module is used to continuously collect real-time vibration data of the corresponding metal structure in the high corrosion risk area of ​​the target offshore converter station, and obtain a vibration time sequence. The high corrosion risk area is determined by analyzing the erosion process of the metal structure in the stress concentration area of ​​the target offshore converter station under the influence of the marine environment. The stress concentration area is determined by analyzing the vibration of the metal structure at different equipment tuning frequencies. The crack data acquisition module is used to input the vibration time series into a preset crack initiation prediction model to generate crack initiation distribution data, and predict crack propagation data based on the crack initiation distribution data; wherein, the crack initiation prediction model is constructed based on the temporal correlation between vibration changes and corrosion rates and Paris's law; The mitigation scheme acquisition module is used to determine the intensity distribution of protective measures based on the crack propagation data, and to acquire the current environmental parameters of the target offshore converter station, so as to determine the corrosion mitigation scheme for the metal structure of the target offshore converter station based on the intensity distribution of protective measures and the current environmental parameters.

9. A terminal device, characterized in that, include: The processor, memory, communication interface, and communication bus are provided, wherein the processor, memory, and communication interface communicate with each other via the communication bus. The memory is used to store at least one executable instruction that causes the processor to perform the operation of the corrosion mitigation method for the metal structure of the offshore converter station as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein, when the computer program is executed, it controls the device or apparatus containing the computer-readable storage medium to perform the corrosion mitigation method for the metal structure of an offshore converter station as described in any one of claims 1 to 7.

Images