Method, device and storage medium for determining the health state of an offshore wind power support structure

Abstract

This application proposes a method, apparatus, and storage medium for determining the health status of offshore wind power support structures. The method includes: acquiring multi-source heterogeneous data of the offshore wind power support structure; performing spatiotemporal alignment and data fusion on the multi-source heterogeneous data to obtain a multi-source synchronous dataset; determining the dynamic corrosion equivalent index based on environmental data in the multi-source synchronous dataset; determining the dynamic corrosion rate based on the base corrosion rate, the dynamic corrosion equivalent index, and the coating degradation factor; calibrating the dynamic corrosion rate based on Bayesian filtering; updating the time-varying corrosion depth based on the dynamic corrosion rate; determining the cumulative corrosion and fatigue damage values ​​of the offshore wind power support structure based on the time-varying corrosion depth and stress cycle information; and determining and outputting the health status level of the offshore wind power support structure based on the cumulative corrosion and fatigue damage values ​​and a preset damage threshold range. This application integrates multi-source data modeling to achieve accurate assessment and intelligent operation and maintenance.

Description

Technical Field

[0001] This invention relates to the field of offshore wind power structural health monitoring technology, and more specifically, to a method, apparatus, and storage medium for determining the health status of offshore wind power support structures. Background Technology

[0002] In related technologies, corrosion assessment of offshore wind power support structures often adopts static empirical models, which cannot respond to dynamic changes in the marine environment. The structural monitoring data is disconnected from the corrosion physical model, resulting in low accuracy and poor real-time performance in health status assessment, which are pain points in the industry. Summary of the Invention

[0003] The present invention aims to solve at least one of the technical problems existing in the prior art or related art.

[0004] Therefore, the first aspect of the present invention proposes a method for determining the health status of offshore wind power support structures.

[0005] A second aspect of the present invention provides a device for determining the health status of offshore wind power support structures.

[0006] A third aspect of the present invention provides another device for determining the health status of offshore wind power support structures.

[0007] The fourth aspect of this application proposes a readable storage medium.

[0008] In view of this, a first aspect of the present invention provides a method for determining the health status of an offshore wind power support structure, comprising: acquiring multi-source heterogeneous data of the offshore wind power support structure, the multi-source heterogeneous data including environmental data, structural monitoring data, and structural foundation data; performing spatiotemporal alignment and data fusion on the multi-source heterogeneous data to obtain a multi-source synchronous dataset; determining the dynamic corrosion equivalent index of the offshore wind power support structure at each monitoring time based on the environmental data in the multi-source synchronous dataset; and determining the health status of the offshore wind power support structure based on the foundation corrosion rate, dynamic corrosion equivalent index in the structural foundation data, and the coating degradation factor at the corresponding monitoring time determined by the structural monitoring data. The system calculates the dynamic corrosion rate at each monitoring time; it then calculates the time-varying corrosion depth based on the cumulative dynamic corrosion rate; it determines the observations based on the structural monitoring data, uses these observations as input for a Bayesian filter, and calibrates the dynamic corrosion rate using the Bayesian filter to obtain the calibrated dynamic corrosion rate; it updates the time-varying corrosion depth based on the calibrated dynamic corrosion rate; it determines the cumulative corrosion and fatigue damage values ​​of the offshore wind power support structure based on the updated time-varying corrosion depth and the stress cycle information represented by the structural monitoring data; and it determines and outputs the health status level of the offshore wind power support structure based on the correspondence between the cumulative corrosion and fatigue damage values ​​and the preset damage threshold range.

[0009] In some technical solutions of this application, environmental data includes time-series data on salinity, temperature, relative humidity, chloride ion concentration, and pH of the sea area where the offshore wind power support structure is located; structural monitoring data includes vibration modal data, strain time history data, natural frequency data, and coating surface image data of the offshore wind power support structure; structural foundation data includes design parameters, material properties, historical inspection records, and factory anti-corrosion scheme data of the offshore wind power support structure, wherein the material properties include the basic corrosion rate under standard conditions.

[0010] In some technical solutions of this application, spatiotemporal alignment and data fusion of multi-source heterogeneous data are performed, including: aligning different data sources on time axes and matching their spatial locations according to the sampling frequency and spatiotemporal reference of the multi-source heterogeneous data; removing outliers from the data according to the historical normal fluctuation range corresponding to the multi-source heterogeneous data, and filling in missing values ​​using linear interpolation or historical mean method; and normalizing the data according to the magnitude and dimension of the filled data to obtain a multi-source synchronous dataset with a unified spatiotemporal reference.

[0011] In some technical solutions of this application, the dynamic corrosion equivalent index is determined as follows: based on the centralized normalized salinity, temperature, relative humidity and chloride ion concentration values ​​of the multi-source synchronous data, combined with the weighting coefficients of the corresponding environmental factors and the material constants of the steel in the structural basis data, the dynamic corrosion equivalent index at the corresponding monitoring time is determined by weighted fitting.

[0012] In some technical solutions of this application, the dynamic corrosion rate is determined by: based on the basic corrosion rate of steel in a standard environment in the structural foundation data, the dynamic corrosion equivalent index at the corresponding monitoring time, and the coating degradation factor at the corresponding monitoring time determined by the structural monitoring data, the dynamic corrosion rate of the offshore wind power support structure at the corresponding time is determined by product fitting.

[0013] In some technical solutions of this application, the coating degradation factor is determined as follows: from the coating surface image data in the structural monitoring data, at least one image feature among crack length, blister area ratio and rust area ratio is extracted; the degree of damage and deterioration of the coating is determined according to the extracted image features, and then the coating degradation factor at the corresponding time is determined; wherein, when the coating is intact and undamaged, the coating degradation factor takes a preset minimum value, and when the coating is completely failed, the coating degradation factor takes a maximum value of 1.

[0014] In some technical solutions of this application, the dynamic corrosion rate is calibrated based on Bayesian filtering, including: establishing a prior distribution of Bayesian filtering based on the predicted damage state characterized by the dynamic corrosion rate and time-varying corrosion depth; constructing observations based on the changes in natural frequency, vibration modes, and strain response in the structural monitoring data, and establishing the likelihood function corresponding to the observations; determining the posterior probability distribution that fits the actual degradation state of the structure based on the Bayesian filtering principle, combined with the prior distribution and the likelihood function; correcting the dynamic corrosion rate based on the posterior probability distribution, and updating the time-varying corrosion depth based on the corrected dynamic corrosion rate.

[0015] A second aspect of the present invention provides a health status determination device for an offshore wind power support structure, comprising: a first acquisition module, a data fusion module, a corrosion index calculation module, a corrosion rate determination module, a corrosion depth calculation module, a Bayesian calibration module, a corrosion depth update module, a damage assessment module, and a status output module. The first acquisition module acquires multi-source heterogeneous data of the offshore wind power support structure, including environmental data, structural monitoring data, and structural foundation data. The data fusion module performs spatiotemporal alignment and data fusion on the multi-source heterogeneous data to obtain a multi-source synchronous dataset. The corrosion index calculation module determines the dynamic corrosion equivalent index of the offshore wind power support structure at each monitoring time based on the environmental data in the multi-source synchronous dataset. The corrosion rate determination module determines the corrosion rate based on the basic corrosion rate, dynamic corrosion equivalent index, and structural monitoring data in the structural foundation data. The system uses the following modules to determine the coating degradation factor at each monitoring time based on the measured data, thus determining the dynamic corrosion rate of the offshore wind power support structure at each monitoring time. The corrosion depth calculation module calculates the time-varying corrosion depth based on the cumulative dynamic corrosion rate. The Bayesian calibration module determines the observations based on the structural monitoring data, uses these observations as input for Bayesian filtering, and calibrates the dynamic corrosion rate based on the Bayesian filter to obtain the calibrated dynamic corrosion rate. The corrosion depth update module updates the time-varying corrosion depth based on the calibrated dynamic corrosion rate. The damage assessment module determines the cumulative corrosion and fatigue damage values ​​of the offshore wind power support structure based on the updated time-varying corrosion depth and the stress cycle information represented by the structural monitoring data. The status output module determines and outputs the health status level of the offshore wind power support structure based on the correspondence between the cumulative corrosion and fatigue damage values ​​and the preset damage threshold range.

[0016] A third aspect of the present invention provides a device for determining the health status of an offshore wind power support structure, comprising: a processor and a memory, wherein the memory stores a program or instructions, and the processor, when executing the program or instructions in the memory, implements the steps of the method for determining the health status of an offshore wind power support structure as described in any of the above-described technical solutions. Therefore, the device for determining the health status of an offshore wind power support structure possesses all the beneficial effects of the method for determining the health status of an offshore wind power support structure as described in any of the above-described technical solutions.

[0017] A fourth aspect of the present invention provides a readable storage medium storing a program or instructions, which, when executed by a processor, implement the steps of the method for determining the health status of an offshore wind power support structure as described in any of the above-described technical solutions. Therefore, the readable storage medium possesses all the beneficial effects of the method for determining the health status of an offshore wind power support structure as described in any of the above-described technical solutions.

[0018] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0019] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which:

[0020] Figure 1 This is one of the flowcharts illustrating a method for determining the health status of an offshore wind power support structure according to an embodiment of the present invention.

[0021] Figure 2 This is a system structure diagram of an offshore wind power support structure life prediction and maintenance optimization system according to an embodiment of the present invention;

[0022] Figure 3 This is a second flowchart illustrating a method for determining the health status of an offshore wind power support structure according to an embodiment of the present invention.

[0023] Figure 4 One of the schematic block diagrams of a health status determination device for offshore wind power support structures according to an embodiment of the present invention;

[0024] Figure 5 A second schematic block diagram of a device for determining the health status of offshore wind power support structures according to an embodiment of the present invention;

[0025] Figure 6 This is one of the interface diagrams of an offshore wind power support structure life prediction and maintenance optimization system according to an embodiment of the present invention;

[0026] Figure 7 This is a second schematic diagram of the interface of an offshore wind power support structure life prediction and maintenance optimization system according to an embodiment of the present invention. Detailed Implementation

[0027] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0028] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0029] The following reference Figures 1 to 7 A method, apparatus, and readable storage medium for determining the health status of offshore wind power support structures are described according to some embodiments of the present invention.

[0030] like Figure 1 As shown in the figure, an embodiment of this application provides a method for determining the health status of an offshore wind power support structure, the steps of which include:

[0031] Step 102: Obtain multi-source heterogeneous data of the offshore wind power support structure. The multi-source heterogeneous data includes environmental data, structural monitoring data, and structural foundation data.

[0032] Step 104: Perform spatiotemporal alignment and data fusion on the multi-source heterogeneous data to obtain a multi-source synchronized dataset;

[0033] Step 106: Determine the dynamic corrosion equivalent index of the offshore wind power support structure at each monitoring time based on the environmental data in the multi-source synchronous dataset.

[0034] Step 108: Determine the dynamic corrosion rate of the offshore wind power support structure at each monitoring time based on the basic corrosion rate and dynamic corrosion equivalent index in the structural foundation data, as well as the coating degradation factor at the corresponding monitoring time determined by the structural monitoring data.

[0035] Step 110: Calculate the time-varying corrosion depth based on the cumulative dynamic corrosion rate;

[0036] Step 112: Determine the observations based on the structural monitoring data, use the observations as input for Bayesian filtering, calibrate the dynamic corrosion rate based on Bayesian filtering, and obtain the calibrated dynamic corrosion rate.

[0037] Step 114: Update the time-varying corrosion depth based on the calibrated dynamic corrosion rate;

[0038] Step 116: Determine the cumulative corrosion and fatigue damage values ​​of the offshore wind power support structure based on the updated time-varying corrosion depth and stress cycle information characterized by structural monitoring data.

[0039] Step 118: Determine and output the health status level of the offshore wind power support structure based on the correspondence between the cumulative damage value of corrosion and fatigue and the preset damage threshold range.

[0040] This application first constructs a refined dynamic corrosion modeling path based on "environmental data, dynamic corrosion equivalent index, and dynamic corrosion rate," avoiding static corrosion models that rely on fixed environmental parameters and empirical formulas. Through spatiotemporal alignment and data fusion of multi-source heterogeneous data, it transforms scattered and asynchronous environmental data, structural monitoring data, and structural basic data into a standardized dataset under a unified spatiotemporal benchmark, laying a high-quality data foundation for subsequent full-process calculations. Then, it calculates the dynamic corrosion equivalent index through environmental data, transforming discrete marine environmental parameters into a comprehensive index that can quantify environmental corrosion and erosion capabilities. This enables the corrosion model to respond in real time to dynamic changes in environmental parameters such as salinity, temperature, and chloride ion concentration in the sea area, fundamentally solving the problem of basic biases in life prediction and damage assessment caused by static corrosion models. Secondly, it innovatively established a Bayesian filter closed-loop calibration mechanism driven by structural monitoring data. It clearly defines the observations extracted from structural monitoring data as the input of Bayesian filtering, uses the corrosion state predicted by the model as prior information, and uses the measured structural response reflecting the actual stiffness degradation and stress changes as observational evidence. Through Bayesian probability statistics, it achieves online calibration of dynamic corrosion rate and synchronous update of time-varying corrosion depth, which solves the industry pain point that the model prediction results continuously deviate from the actual state of the structure over the service time, and enables the corrosion damage calculation results to track the actual degradation process of the structure in real time. Furthermore, it constructs a corrosion-fatigue coupled cumulative damage assessment system. Based on the calibrated time-varying corrosion depth and stress cycle information characterized by structural monitoring data, it comprehensively calculates the corrosion and fatigue cumulative damage values ​​of the structure. This fully considers the accelerating effect of corrosion-induced cross-sectional weakening on fatigue damage, avoiding the unsafe results caused by evaluating a single damage. Finally, through the correspondence between the cumulative damage value and a preset damage threshold range, it completes the quantitative determination and engineering output of the health status level. Based on the above core invention points, this application realizes a complete technical chain from multi-source data standardization processing, dynamic corrosion modeling, measured data closed-loop calibration, coupled damage quantitative assessment to health status level determination. The ring fundamentally overcomes the three core defects of static models, data and model separation, and incomplete damage assessment. The core technical effect is reflected in not only significantly improving the accuracy, real-time performance, and reliability of corrosion rate prediction and damage assessment of offshore wind power support structures, but also constructing a set of feasible, repeatable, and verifiable technical solutions for determining the structural health status. This provides a highly confident quantitative basis for subsequent operation and maintenance decisions of offshore wind power support structures, fills the technical gap of insufficient deep integration of dynamic corrosion modeling and online closed-loop calibration of offshore wind power support structures, and realizes a fundamental improvement in the assessment of the health status of offshore wind power support structures from static, offline, and experience-oriented to dynamic, online, and data and physical model integration.

[0041] Among them, the offshore wind power support structure refers to the substructure of the offshore wind turbine that carries the upper structure such as the nacelle, tower, and blades and transfers the upper load to the seabed foundation. It includes foundation forms such as jackets, monopiles, suction buckets, and tripods, as well as the lower support section of the tower. It is the core load-bearing component that directly bears the corrosion and alternating loads of the marine environment, and is also the direct target of this technical solution.

[0042] Multi-source heterogeneous data refers to a multi-dimensional data set related to the service status of offshore wind power support structures, which comes from different acquisition terminals, different data types, different sampling frequencies, and different spatiotemporal references. It is the basic input of this scheme and is specifically divided into three categories: environmental data, structural monitoring data, and structural basic data.

[0043] Environmental data refers to time-series data of various environmental parameters that can reflect the corrosion and erosion capabilities of the marine environment in which the offshore wind power support structure is located. It is the core input that drives the dynamic updating of the corrosion model.

[0044] Structural monitoring data refers to real-time monitoring data obtained directly from various sensors and acquisition devices deployed on offshore wind power support structures. This data reflects changes in the structure's mechanical properties, material state, and surface protection status, and serves as the core basis for model calibration and damage calculation.

[0045] Structural basic data refers to the basic attribute data of offshore wind power support structures that remain relatively fixed throughout their entire life cycle. It serves as the basic benchmark for constructing corrosion models and determining damage thresholds.

[0046] Spatiotemporal alignment refers to unifying the time dimension benchmark and matching the spatial dimension position of multi-source heterogeneous data from different sources, with different sampling frequencies and different collection locations, thereby eliminating the time difference and spatial location deviation between different data sources.

[0047] Data fusion refers to the standardization process of multi-source data that has been aligned in time and space, including outlier removal, missing value completion, and dimensional unification, to eliminate the interference of data noise and dimensional differences on subsequent calculations.

[0048] Multi-source synchronous datasets refer to standardized multi-dimensional data sets that have undergone spatiotemporal alignment and data fusion processing, are under a unified spatiotemporal benchmark, and whose data quality meets the requirements for subsequent calculations. They serve as direct inputs for subsequent dynamic corrosion modeling and damage calculation.

[0049] The Dynamic Corrosion Equivalent Index (DCEI) is a comprehensive quantitative indicator proposed in this scheme to quantify the corrosion and erosion capacity of steel used in offshore wind power support structures by the target marine environment. It can respond in real time to the spatiotemporal dynamic changes of marine environmental parameters and is a core intermediate parameter connecting environmental data and dynamic corrosion rate.

[0050] The base corrosion rate refers to the uniform corrosion rate of the steel used in offshore wind power support structures under standard laboratory conditions, without coating protection, and with fixed environmental parameters. It is the benchmark value for calculating the dynamic corrosion rate under actual service conditions and is derived from the material property parameters in the structural foundation data.

[0051] The coating degradation factor is a dimensionless parameter used to quantify the degree of degradation of the protective performance of the anti-corrosion coating on the surface of offshore wind power support structures. It can reflect the changes in the coating's ability to inhibit the corrosion process of steel and is a key parameter for correcting dynamic corrosion rates.

[0052] Dynamic corrosion rate refers to the increase in corrosion thickness of steel per unit time in the actual service environment of offshore wind power support structures, determined by combining real-time environmental erosion capacity, coating protection status, and the corrosion characteristics of the materials themselves. It is a core dynamic parameter reflecting the corrosion process of the structure.

[0053] Time-varying corrosion depth refers to the total corrosion thickness of the steel in an offshore wind power support structure, calculated from the initial service time to the corresponding monitoring time based on the cumulative dynamic corrosion rate over time. It is a core parameter characterizing the degree of structural corrosion damage.

[0054] Observations refer to measured physical quantities that can be directly obtained from structural monitoring data during the Bayesian filtering process and have a clear physical correspondence with the actual damage state of the structure. They are the core input for achieving the fusion and calibration of the model's predicted state and the measured state.

[0055] Bayesian filtering is a recursive state estimation algorithm that combines the prior state predicted by the model with the actual observation data measured on site, based on the Bayesian probability and statistics principle, to solve the posterior probability distribution of the true state of the structure. It is the core technical means for realizing online calibration of the corrosion model in this scheme.

[0056] The calibrated dynamic corrosion rate refers to the dynamic corrosion rate that, after Bayesian filtering calibration and correction, better reflects the actual corrosion process of the offshore wind power support structure, eliminating the deviation between the static model prediction and the actual state of the structure.

[0057] Stress cycle information refers to parameters such as the number of cycles and stress amplitude of alternating stress generated by offshore wind power support structures under wind load, wave load, and unit operating load, extracted from the strain time history data of structural monitoring data. It is the core basis for calculating fatigue damage.

[0058] The cumulative damage value of corrosion and fatigue refers to the total cumulative damage degree of the structure obtained by taking into account the cross-sectional weakening damage caused by uniform corrosion of the offshore wind power support structure and the fatigue damage caused by alternating loads. It is a core quantitative indicator that characterizes the overall health status of the structure.

[0059] The preset damage threshold range refers to the cumulative damage value range corresponding to different health status levels, which is pre-divided according to the design parameters, material properties, and safety specifications of the offshore wind power support structure. It serves as a quantitative benchmark for determining the health status level of the structure.

[0060] The health status level refers to the classification of the overall service health status of the offshore wind power support structure based on the preset damage threshold range of the cumulative damage values ​​of corrosion and fatigue. It is the final output of this technical solution.

[0061] In some embodiments of this application, environmental data includes time-series data on salinity, temperature, relative humidity, chloride ion concentration, and pH of the sea area where the offshore wind power support structure is located; structural monitoring data includes vibration modal data, strain time history data, natural frequency data, and coating surface image data of the offshore wind power support structure; and structural foundation data includes design parameters, material properties, historical inspection records, and factory anti-corrosion scheme data of the offshore wind power support structure, wherein material properties include the basic corrosion rate under standard conditions.

[0062] In the above embodiments, the salinity time-series data is a series of salinity values ​​in the seawater or salt spray of the area where the offshore wind power support structure is located, collected continuously at fixed time intervals. The temperature time-series data is a series of seawater temperature and ambient air temperature values ​​in the area where the offshore wind power support structure is located, collected continuously at fixed time intervals. The relative humidity time-series data is a series of relative humidity values ​​in the atmospheric environment where the offshore wind power support structure is located, collected continuously at fixed time intervals. The chloride ion concentration time-series data is a series of chloride ion concentration values ​​in the seawater or salt spray where the offshore wind power support structure is located, collected continuously at fixed time intervals. The pH (acidity / alkalinity) value series of the seawater where the offshore wind power support structure is located is collected continuously at fixed time intervals. These environmental data collectively constitute the core basic parameters that can reflect the corrosion and erosion resistance of the marine environment where the offshore wind power support structure is located. The vibration modal data are the dynamic characteristic parameters of the offshore wind power support structure under inherent excitation, such as mode shape, damping ratio, and modal frequency, collected by vibration sensors. Strain time history data is a continuous sequence of strain values ​​at a specified location on the offshore wind turbine support structure under load, collected by strain gauges, over time. Natural frequency data represents the natural vibration frequencies of the offshore wind turbine support structure. Coating surface image data consists of high-resolution images of the anti-corrosion coating on the surface of the offshore wind turbine support structure, captured by an image acquisition device. This structural monitoring data directly reflects changes in the mechanical properties, material state, and surface protection status of the offshore wind turbine support structure. Design parameters are the fundamental design parameters specified in the offshore wind turbine support structure design drawings, including structural dimensions, cross-sectional parameters, design load, design life, and safety factor. Material properties include the grade, yield strength, elastic modulus, corrosion resistance, and fatigue performance of the steel used in the offshore wind turbine support structure. Historical inspection records are historical data records generated from various offline inspections, underwater inspections, and anti-corrosion maintenance during the service life of the offshore wind turbine support structure. Factory anti-corrosion scheme data is the basic data regarding the type, thickness, anti-corrosion process, and design protection life of the anti-corrosion coating used when the offshore wind turbine support structure leaves the factory. The baseline corrosion rate under standard conditions is the uniform corrosion rate of marine steel measured under standard marine environmental laboratory conditions, without coating protection, with fixed environmental parameters, and under stable corrosion conditions.These structural foundational data provide a fixed benchmark and historical reference for corrosion model construction, damage threshold determination, and condition assessment. These clearly categorized, multi-source heterogeneous data comprehensively cover all core dimensions affecting the corrosion process, mechanical performance, and health status of offshore wind power support structures. They provide clear and directly obtainable data sources for all subsequent steps, including dynamic corrosion equivalent index calculation, coating degradation factor calculation, Bayesian calibration, and damage calculation, forming a complete data source closed loop. This ensures that all input parameters of the health status determination method have stable and reliable data sources. At the same time, the specific composition of various types of data is clearly defined, avoiding the problem of ambiguous protection scope caused by overly broad data coverage, and ensuring the repeatability and feasibility of the technical solution.

[0063] For example, on the monopile support structure of a 3.6MW offshore wind turbine in a certain offshore wind farm, marine environmental sensors deployed around the monopile foundation collect time-series data on salinity, temperature, relative humidity, chloride ion concentration, and pH at a sampling frequency of once per hour. Vibration sensors and strain gauges deployed on the monopile tower collect vibration modal data, strain time history data, and natural frequency data at a sampling frequency of 100Hz. High-definition image acquisition devices deployed on key parts of the tower acquire coating surface image data at a sampling frequency of once per day. At the same time, the design parameters of the monopile support structure, the material properties of the first marine engineering steel used, historical underwater inspection records during service, and factory heavy corrosion protection scheme data are retrieved from the wind farm asset operation and maintenance management system. Among them, the material properties include the basic corrosion rate of the first marine engineering steel under standard environment, thus obtaining all-dimensional basic data required for the health status determination method.

[0064] In some embodiments of this application, spatiotemporal alignment and data fusion of multi-source heterogeneous data include: aligning different data sources on time axes and matching their spatial locations based on the sampling frequency and spatiotemporal reference of the multi-source heterogeneous data; removing outliers from the data based on the historical normal fluctuation range corresponding to the multi-source heterogeneous data, and filling in missing values ​​using linear interpolation or historical mean methods; and normalizing the data based on the magnitude and dimensions of the filled data to obtain a multi-source synchronized dataset with a unified spatiotemporal reference.

[0065] In the above embodiments, the sampling frequency is the number of times various sensors collect data per unit time. The spatiotemporal reference is the time coordinate system and spatial coordinate system used for data acquisition. Time axis alignment is the process of mapping data sources with different sampling frequencies onto the same time axis, and synchronizing data from different data sources in the time dimension through resampling, interpolation, and other methods. Spatial location matching is the process of binding the collected data with the corresponding parts of the offshore wind power support structure according to the actual installation location of the sensor, thus clarifying the specific structural location corresponding to the monitoring data.

[0066] The historical normal fluctuation range is the range of normal numerical fluctuations determined by statistical methods based on historical data from the corresponding data source during the service life of the offshore wind power support structure under normal conditions of no anomalies and no damage. Outliers are abnormal data points that exceed the historical normal fluctuation range and do not conform to the normal service state of the structure and the data collection patterns. Linear interpolation is a method of supplementing missing data values ​​within a time period by fitting a linear function between two adjacent valid data points. The historical mean method is a method of supplementing missing data values ​​for time periods with missing values ​​by using the historical normal mean of the corresponding time period from the data source.

[0067] Normalization is a process that maps data of different magnitudes and units to a unified numerical range through linear transformations and other methods, eliminating differences in units and magnitudes between different parameters. A unified spatiotemporal benchmark multi-source synchronized dataset is a standardized multi-dimensional data set where all data are in the same time coordinate system, the same spatial coordinate system, and the same numerical unit after spatiotemporal alignment and data fusion.

[0068] By aligning the time axis and matching the spatial location, a unified benchmark for multi-source heterogeneous data was achieved from both temporal and spatial dimensions. This resolved the spatiotemporal misalignment of data caused by differences in sampling frequencies and installation locations of different sensors, ensuring accurate matching of environmental data, structural monitoring data, and corresponding structural parts in subsequent calculations. Outlier removal and missing value completion effectively eliminated interference from data noise and missing data in subsequent calculations, improving the quality and reliability of the input data. Normalization eliminated dimensional and magnitude differences between different environmental parameters, ensuring that the weights of each environmental factor in the subsequent dynamic corrosion equivalent index weighted fitting calculation accurately reflect their impact on the corrosion process, avoiding weight distortion caused by differences in parameter magnitudes.

[0069] The resulting unified spatiotemporal benchmark multi-source synchronous dataset provides a high-quality and highly reliable input foundation for subsequent dynamic corrosion modeling and damage calculation, further improving the accuracy of health status assessment results from the data processing level.

[0070] For example, for multi-source heterogeneous data collected from the monopile support structure of an offshore wind turbine, Beijing time was first used as the unified time reference. Strain time history data and vibration data with a sampling frequency of 100Hz were downsampled and resampled to align with environmental data with a sampling frequency of 1 time per hour on the same hourly time axis. Simultaneously, based on the sensor installation location, strain and vibration data at the bottom of the tower were bound to the corrosion calculation location of the monopile foundation, completing time axis alignment and spatial location matching. Then, based on the historical data of the wind farm over the past three years, the historical normal fluctuation ranges of various environmental parameters and structural monitoring parameters were determined, and abnormal data points exceeding the upper and lower limits of the range were removed. For missing data due to sensor communication interruptions, environmental parameters were supplemented using the historical average of the corresponding time period, and strain time history data were supplemented using linear interpolation. Finally, a minimum-maximum normalization method was used to map all environmental parameters of different dimensions, such as salinity, temperature, relative humidity, and chloride ion concentration, to the numerical range [0,1], ultimately obtaining a multi-source synchronous dataset under a unified spatiotemporal reference, completing the entire data fusion process.

[0071] In some embodiments of this application, the dynamic corrosion equivalent index is determined as follows: based on the centralized normalized salinity, temperature, relative humidity and chloride ion concentration values ​​of the multi-source synchronous data, combined with the weighting coefficients of the corresponding environmental factors and the material constants of the steel in the structural basis data, the dynamic corrosion equivalent index at the corresponding monitoring time is determined by weighted fitting.

[0072] In the above embodiments, the normalized salinity, temperature, relative humidity, and chloride ion concentration values ​​are environmental parameters mapped to a unified numerical range after normalization. The weighting coefficients of the environmental factors are weighted coefficients used to characterize the degree of influence of each environmental parameter on the steel corrosion process. The material constants of the steel are fixed constants related to the material grade and corrosion characteristics of the steel used in the offshore wind power support structure.

[0073] Weighted fitting is a method for calculating the dynamic corrosion equivalent index by using normalized environmental parameters as independent variables and the weighting coefficients of the corresponding environmental factors as weighting values, and then fitting the data through linear weighted summation. The dynamic corrosion equivalent index is a comprehensive quantitative indicator used to quantify the corrosion and erosion capabilities of the target marine environment on the steel used in offshore wind power support structures.

[0074] By employing a weighted fitting method, discrete environmental parameters of different dimensions are transformed into a single quantitative index that comprehensively reflects the environmental corrosion and erosion capacity. Weighting coefficients differentiate the degree of influence of different environmental factors on the corrosion process, enabling the dynamic corrosion equivalent index to accurately reflect the environmental corrosion and erosion capacity of the target sea area. Material constants are used to accommodate the varying corrosion resistance of different marine engineering steels, making the calculation of the dynamic corrosion equivalent index applicable to different types of offshore wind power support structures, thus improving the versatility and adaptability of the technical solution.

[0075] Meanwhile, the clearly defined weighted fitting calculation method is entirely based on directly collectable environmental data and known material parameters, without any black-box calculation process, ensuring the repeatability and feasibility of the technical solution. It fundamentally avoids the inherent limitation of static empirical corrosion models failing to respond to dynamic environmental changes, achieving dynamic and accurate quantification of corrosion erosion capacity, and providing core driving parameters for the accurate calculation of dynamic corrosion rates.

[0076] For example, for Q355ND marine steel used in the monopile support structure of an offshore wind farm, the following steps were taken: first, historical corrosion test data and orthogonal experiments were used to determine the weighting coefficients for salinity (0.42), temperature (0.25), relative humidity (0.13), and chloride ion concentration (0.20). Based on the corrosion performance of Q355ND steel, the material constant was determined to be 0.05. Then, normalized salinity (0.68), temperature (0.72), relative humidity (0.55), and chloride ion concentration (0.81) values ​​were extracted from the multi-source synchronous dataset. Finally, the dynamic corrosion equivalent index at the corresponding monitoring time was calculated to be 1.0021 through weighted fitting.

[0077] In some embodiments of this application, the dynamic corrosion rate is determined by: based on the basic corrosion rate of steel under standard environment in the structural foundation data, the dynamic corrosion equivalent index at the corresponding monitoring time, and the coating degradation factor at the corresponding monitoring time determined by the structural monitoring data, the dynamic corrosion rate of the offshore wind power support structure at the corresponding time is determined by product fitting.

[0078] In the above embodiments, the basic corrosion rate of steel under standard conditions is the uniform corrosion rate of marine engineering steel measured under standard marine environmental laboratory conditions, without coating protection, with fixed environmental parameters, and in a stable corrosion state. The dynamic corrosion equivalent index corresponding to the monitoring time is a comprehensive quantitative indicator of the corrosion and erosion capacity of the target sea area environment at the corresponding monitoring time, calculated through weighted fitting. The coating degradation factor corresponding to the monitoring time is a dimensionless parameter that can quantify the degree of deterioration of the protective performance of the anti-corrosion coating on the surface of the offshore wind power support structure at the corresponding monitoring time.

[0079] Product fitting is a method for calculating the dynamic corrosion rate under actual service conditions by multiplying three parameters: the basic corrosion rate, the dynamic corrosion equivalent index, and the coating degradation factor. The dynamic corrosion rate is the increase in steel corrosion thickness per unit time, determined by the offshore wind turbine support structure under actual service conditions, taking into account real-time environmental erosion capabilities, coating protection status, and the material's own corrosion characteristics.

[0080] By employing a product-fitting calculation method, this approach achieves coupled calculations of three core dimensions: material baseline corrosion characteristics, real-time environmental erosion capability, and coating protection status. This enables the dynamic corrosion rate to comprehensively and accurately reflect the actual corrosion process of the structure in real-world service environments. Unlike existing corrosion rate calculation methods that only consider environmental factors, this approach introduces a coating degradation factor as a core correction term. This accurately reflects the accelerating effect of anti-corrosion coating deterioration on the corrosion rate, fundamentally improving the accuracy of dynamic corrosion rate calculations.

[0081] This process provides reliable core parameters for subsequent corrosion depth accumulation calculation and damage assessment. Simultaneously, the clearly defined product fitting calculation method forms a complete closed loop for dynamic corrosion rate calculation, ensuring the feasibility and repeatability of the technical solution.

[0082] For example, for a monopile support structure of an offshore wind turbine, the corrosion rate of Q355ND marine steel under standard conditions is first obtained from the structural foundation data as 0.08 mm / year. The dynamic corrosion equivalent index at the corresponding monitoring time is obtained as 1.0021 from the calculated results of the dynamic corrosion equivalent index. The coating degradation factor at the corresponding monitoring time is obtained as 0.32 from the calculated results of the structural monitoring data. Finally, the dynamic corrosion rate at the corresponding monitoring time is calculated as 0.02565 mm / year through product fitting.

[0083] In some embodiments of this application, the coating degradation factor is determined as follows: from the coating surface image data in the structural monitoring data, at least one image feature is extracted, including crack length, blister area ratio, and rust area ratio; the degree of damage and deterioration of the coating is determined based on the extracted image features, and then the coating degradation factor at the corresponding time is determined; wherein, when the coating is intact and undamaged, the coating degradation factor takes a preset minimum value, and when the coating is completely failed, the coating degradation factor takes a maximum value of 1.

[0084] In the above embodiments, the coating surface image data is high-resolution image data of the anti-corrosion coating on the surface of the offshore wind power support structure, captured by an image acquisition device. Image features are extracted from the coating surface image data and are visual feature parameters that quantify the damage and deterioration state of the coating. Crack length is the sum of the linear lengths of cracks appearing on the anti-corrosion coating surface. Blister area percentage is the proportion of the area of ​​blistering and peeling defects on the coating surface to the total inspected area of ​​the coating. Rust area percentage is the proportion of the area of ​​steel substrate rust on the coating surface to the total inspected area of ​​the coating.

[0085] The degree of damage and degradation is a quantitative measure of the deterioration of the coating's protective performance, derived from a comprehensive analysis of defects such as cracks, blistering, and corrosion. The preset minimum value represents the coating degradation factor when the coating is intact and undamaged, corresponding to the optimal state of coating protective performance. The coating degradation factor is a dimensionless parameter that quantifies the degree of degradation of the anti-corrosion coating's protective performance on the surface of offshore wind turbine support structures.

[0086] By extracting quantifiable and verifiable engineering image features from coating surface image data, the abstract coating state recognition process is transformed into a concrete feature extraction and quantification calculation process. This approach covers all image recognition implementation methods while anchoring to specific engineering physical characteristics, thus avoiding the identification risks associated with purely algorithmic solutions.

[0087] By determining the degree of coating damage and degradation through image features, and further determining the coating degradation factor, a complete calculation chain from image data to the coating degradation factor is formed, ensuring the feasibility of the technical solution. Simultaneously, the boundary values ​​of the coating degradation factor are clearly defined, ensuring complete self-consistency in the calculation logic of fitting the product of the coating degradation factor value and the dynamic corrosion rate. The degradation factor value is smallest when the coating is intact, corresponding to the lowest dynamic corrosion rate. The degradation factor value is largest when the coating is completely failed, corresponding to the highest dynamic corrosion rate, ensuring a logical closed loop in the technical solution and providing a reliable parameter basis for the accurate correction of the dynamic corrosion rate.

[0088] For example, regarding the anti-corrosion coating on the surface of a monopile tower of an offshore wind turbine, high-definition image acquisition devices were used to acquire surface image data of the coating. Three image features were extracted from the images: the total length of coating cracks (12.5 cm), the percentage of blistering area (2.1%), and the percentage of rusted area (0.8%). Based on a preset coating degradation level assessment standard and the extracted three image features, the coating damage and degradation degree in the monitored area was determined to be mild, with a corresponding degradation degree quantification value of 0.32. The preset minimum value is 0 when the coating is intact and undamaged, and the maximum value is 1 when the coating is completely failed. Finally, the coating degradation factor at the corresponding monitoring time was determined to be 0.32.

[0089] In some embodiments of this application, the dynamic corrosion rate is calibrated based on Bayesian filtering, including: establishing a prior distribution of Bayesian filtering based on the predicted damage state characterized by the dynamic corrosion rate and the time-varying corrosion depth; constructing observations based on the changes in natural frequency, vibration modes, and strain response in the structural monitoring data, and establishing the likelihood function corresponding to the observations; determining a posterior probability distribution that fits the actual degradation state of the structure based on the Bayesian filtering principle, combined with the prior distribution and the likelihood function; correcting the dynamic corrosion rate based on the posterior probability distribution, and updating the time-varying corrosion depth based on the corrected dynamic corrosion rate.

[0090] In the above embodiments, the predicted damage state is obtained by calculating the corrosion damage prediction state of the offshore wind power support structure using dynamic corrosion rate and time-varying corrosion depth. The prior distribution is the probability distribution of the structural damage state predicted based on the dynamic corrosion model before the introduction of measured observation data.

[0091] The change in natural frequency is the difference between the natural frequency of the structure at the current monitoring moment and the natural frequency of the structure in its initial, intact state. The change in vibration modes is the difference between the vibration mode parameters of the structure at the current monitoring moment and the vibration mode parameters of the structure in its initial, intact state. The change in strain response is the difference between the strain response of the structure at the current monitoring moment and the strain response of the structure in its initial, intact state under the same load.

[0092] Observations are measured physical quantities that can be directly obtained from structural monitoring data during Bayesian filtering and have a clear physical correspondence with the actual damage state of the structure. The likelihood function, in Bayesian filtering, is a function used to characterize the conditional probability of the corresponding measured value of an observation occurring under a given structural damage state.

[0093] Bayesian filtering is based on Bayesian probability and statistics principles. It combines the model's predicted prior state with actual field observation data to recursively estimate the posterior probability distribution of the structure's true state. The posterior probability distribution is calculated using Bayes' theorem, combining the prior distribution and the likelihood function, to represent the probability distribution of the structure's true damage state.

[0094] The modified dynamic corrosion rate is a process of adjusting the original dynamic corrosion rate based on the posterior probability distribution to better reflect the actual corrosion process of the structure. The updated time-varying corrosion depth is a process of recalculating the total corrosion thickness of the structure based on the modified dynamic corrosion rate.

[0095] By establishing a prior distribution, the prediction results of the corrosion model are transformed into prior information in Bayesian filtering, thus clarifying the initial benchmark for state estimation. By constructing observations including changes in natural frequency, vibration modes, and strain response, and establishing corresponding likelihood functions, measured structural response data that reflects the actual mechanical performance degradation of the structure are transformed into observational evidence in Bayesian filtering.

[0096] This process clarified the specific physical composition of the observations and established a probabilistic correlation between the observation data and the state estimate. By calculating the posterior probability distribution, a deep fusion of model prediction information and field measurement information was achieved, resulting in a damage state estimate that more closely reflects the actual state of the structure.

[0097] By correcting the dynamic corrosion rate and updating the time-varying corrosion depth, the probabilistic estimation results were transformed into physical parameters that can be directly used for damage calculation, thus completing the online closed-loop calibration of the corrosion model. The entire process established a complete closed loop from model prediction to experimental calibration, enabling the dynamic corrosion model to track the actual degradation process of the structure in real time.

[0098] This fundamentally solves the industry pain point of static corrosion models drifting over time, significantly improves the accuracy and reliability of corrosion state prediction, and provides a precise parameter basis for subsequent corrosion-fatigue coupled damage calculation.

[0099] For example, for a monopile support structure of an offshore wind turbine, a normal prior distribution of the corrosion damage state is first established based on the predicted damage state characterized by a dynamic corrosion rate of 0.02565 mm / year and a time-varying corrosion depth of 2.45 mm. The prior mean is 2.45 mm, and the prior variance is 0.12 mm². Then, observations are constructed by extracting the natural frequency change (-0.82 Hz), vibration mode change, and strain response change from the structural monitoring data. Based on the physical mapping relationship between the observations and the corrosion damage state, a corresponding normal likelihood function is established, with a likelihood mean of 2.68 mm and a likelihood variance of 0.08 mm².

[0100] Based on the Bayesian filtering principle, combined with the prior distribution and likelihood function, the posterior probability distribution of the structural corrosion depth was calculated, with a posterior mean of 2.61 mm and a posterior variance of 0.05 mm². Finally, the dynamic corrosion rate was corrected to 0.0278 mm / year based on the mean of the posterior distribution, and the time-varying corrosion depth was updated to 2.61 mm based on the corrected dynamic corrosion rate, completing the entire Bayesian filtering calibration process.

[0101] like Figure 2 As shown, Figure 2This paper showcases the overall architecture of a fully closed-loop intelligent operation and maintenance system for offshore wind power support structures. The system, from top to bottom, comprises a multi-source data acquisition layer, a data fusion and preprocessing layer, a core intelligent analysis and decision engine, a decision output and feedback execution layer, and a model self-learning and optimization module. The multi-source data acquisition layer integrates three main data sources: environmental data sources, SHM monitoring data sources, and operation and maintenance data sources, providing a complete multi-source heterogeneous data input foundation for the offshore wind power support structure health status determination method in this application. The data fusion and preprocessing center, deployed within an edge gateway or site server, completes the standardized processing of multi-source data through spatiotemporal data alignment, feature extraction and fusion, quality checks, and interpolation, providing a high-quality synchronous dataset for the construction of the dynamic environmental corrosion model and online calibration of model parameters. The core intelligent analysis and decision engine, deployed in the cloud or on a high-performance server, sequentially connects a dynamic environmental corrosion model, a Bayesian state updater, a damage evolution predictor, and a maintenance window optimizer, fully corresponding to the connection logic between the core technical process and subsequent operation and maintenance decisions in this application. The dynamic environmental corrosion model implements the DCEI index and dynamic corrosion... The rate calculation and Bayesian state updater complete the reverse calibration of model parameters based on structural monitoring data. The damage evolution predictor realizes the calculation of cumulative damage degree and the probability distribution prediction of remaining service life (RUL). The above three modules together constitute the core technology core for determining the health status of offshore wind power support structures. The maintenance window optimizer, based on the risk characterization parameters and structural failure probability output by this application, completes the optimization and generation of maintenance plans under multiple constraints. The decision output and feedback execution layer realizes the visualization of structural health status, risk warning prompts, implementation of maintenance plans, and full-process feedback of operation and maintenance effect data through three functional units: human-computer interaction interface, operation and maintenance command execution, and effect feedback collection. Finally, the returned structural re-inspection data, actual operation and maintenance cost data, and maintenance performance evaluation data are input into the model self-learning and optimization module, realizing the automatic adjustment of parameters and continuous improvement of prediction accuracy of the dynamic corrosion model and damage prediction algorithm in this application. Finally, a closed-loop operation system is formed from multi-source data acquisition, accurate assessment of structural health status, intelligent operation and maintenance decision generation to operation and maintenance effect feedback and model iterative optimization.

[0102] like Figure 3 As shown, an embodiment of this application provides a method for determining the health status of an offshore wind power support structure, the steps of which include:

[0103] Step 201: Initiate the offshore wind power support structure life prediction and maintenance optimization process;

[0104] Step 202: Perform multi-source data acquisition and fusion processing to obtain a synchronized dataset with a unified spatiotemporal reference;

[0105] Step 203: Construct and update the dynamic environment corrosion model, and calculate the dynamic corrosion equivalent index and dynamic corrosion rate;

[0106] Step 204: Based on the Bayesian update principle, the corrosion model state is calibrated by integrating monitoring data;

[0107] Step 205: Conduct damage evolution simulation and remaining service life prediction, calculate cumulative damage and output the remaining service life probability distribution;

[0108] Step 206: Perform maintenance window optimization under multiple constraints to generate the optimal maintenance time window sequence with the goal of minimizing cost;

[0109] Step 207: Implement maintenance decisions, conduct maintenance effectiveness assessments, and complete model parameter adjustments;

[0110] Step 208: Continuously iterate and update based on feedback, or end the current process.

[0111] like Figure 4As shown, an embodiment of this application provides a health status determination device 300 for offshore wind power support structures, including: a first acquisition module 310, a data fusion module 320, a corrosion index calculation module 330, a corrosion rate determination module 340, a corrosion depth calculation module 350, a Bayesian calibration module 360, a corrosion depth update module 370, a damage assessment module 380, and a status output module 390. The first acquisition module 310 is used to acquire multi-source heterogeneous data of the offshore wind power support structure, including environmental data, structural monitoring data, and structural basic data. The data fusion module 320 is used to perform spatiotemporal alignment and data fusion on the multi-source heterogeneous data to obtain a multi-source synchronous dataset. The corrosion index calculation module 330 is used to determine the dynamic corrosion equivalent index of the offshore wind power support structure at each monitoring time based on the environmental data in the multi-source synchronous dataset. The corrosion rate determination module 340 is used to determine the basic corrosion rate in the structural basic data based on the basic corrosion rate. The system employs a dynamic corrosion equivalent index and a coating degradation factor determined from structural monitoring data at corresponding monitoring times to determine the dynamic corrosion rate of the offshore wind power support structure at each monitoring time. A corrosion depth calculation module 350 calculates the time-varying corrosion depth based on the dynamic corrosion rate. A Bayesian calibration module 360 ​​determines observations from structural monitoring data, uses these observations as input for Bayesian filtering, and calibrates the dynamic corrosion rate based on the Bayesian filter to obtain the calibrated dynamic corrosion rate. A corrosion depth update module 370 updates the time-varying corrosion depth based on the calibrated dynamic corrosion rate. A damage assessment module 380 determines the cumulative corrosion and fatigue damage values ​​of the offshore wind power support structure based on the updated time-varying corrosion depth and the stress cycle information characterized by the structural monitoring data. A status output module 390 determines and outputs the health status level of the offshore wind power support structure based on the correspondence between the cumulative corrosion and fatigue damage values ​​and a preset damage threshold range.

[0112] This application provides a health status determination device 300 for offshore wind power support structures, comprising: a first acquisition module 310, a data fusion module 320, a corrosion index calculation module 330, a corrosion rate determination module 340, a corrosion depth calculation module 350, a Bayesian calibration module 360, a corrosion depth update module 370, a damage assessment module 380, and a status output module 390. Through a modular architecture corresponding one-to-one with the health status determination method for offshore wind power support structures, it transforms the entire process of multi-source heterogeneous data acquisition, spatiotemporal alignment and data fusion, dynamic corrosion equivalent index calculation, dynamic corrosion rate determination, time-varying corrosion depth calculation, Bayesian filtering calibration, corrosion depth update, corrosion and fatigue cumulative damage assessment, and health status level output into physical functional modules. This allows for stable, efficient, and automated execution of the offshore wind power support structure health status determination process. Relying on a complete data processing link and physical model calibration mechanism, it effectively improves the accuracy, real-time performance, and reliability of structural health assessment, providing stable and implementable hardware support for the safe operation and maintenance of offshore wind power support structures.

[0113] like Figure 5 As shown, an embodiment of this application provides a health status determination device 400 for offshore wind power support structures, including a processor 402 and a memory 404. The memory 404 stores programs or instructions. When the processor 402 executes the programs or instructions in the memory 404, it implements the steps of the offshore wind power support structure health status determination method as described in any of the above embodiments. Therefore, the offshore wind power support structure health status determination device 400 possesses all the beneficial effects of the offshore wind power support structure health status determination method as described in any of the above embodiments.

[0114] Embodiments of this application provide a readable storage medium storing a program or instructions that, when executed by a processor, implement the steps of the method for determining the health status of an offshore wind power support structure as described in any of the above embodiments. Therefore, the readable storage medium possesses all the beneficial effects of the method for determining the health status of an offshore wind power support structure as described in any of the above embodiments.

[0115] like Figure 6 and Figure 7 As shown, Figure 6 and Figure 7This is a schematic diagram of the interface for the offshore wind power support structure life prediction and maintenance optimization system. The interface displays an overview of the status of key components, dynamic changes in the environmental corrosion equivalent index, probability distribution of remaining life, suggested sequences of optimized maintenance windows, maintenance cost-benefit analysis, system operation status, risk warning status, data source and model update status, model prediction error trends, and risk evolution trend predictions. It intuitively presents core information such as dynamic corrosion equivalent index, remaining life, failure probability, maintenance time window, operation type, estimated cost, and risk reduction extent. It integrates and displays the dynamic environmental corrosion model, the damage state after Bayesian calibration, probabilistic remaining life prediction, and the optimal maintenance plan, providing operation and maintenance personnel with a visualized and directly executable intelligent operation and maintenance decision-making basis. This embodiment discloses a method for generating a maintenance scheme for offshore wind power support structures. The specific implementation steps are as follows: First, obtain the risk characterization parameters of the offshore wind power support structure. The risk characterization parameters include at least one of the following: structural failure probability, remaining available time, and risk level. Then, obtain the corresponding sea state operation parameters, operation and maintenance resource parameters, and shutdown impact parameters. The sea state operation parameters include wind speed, wave height, and visibility parameters for the corresponding time period in the sea area. The operation and maintenance resource parameters include the available time period parameters for operation and maintenance vessels, personnel, and equipment materials. The shutdown impact parameters include the grid dispatch-allowed shutdown time period parameters and the shutdown loss coefficient. Based on the sea state operation parameters, filter the set of candidate operation time periods that meet the operation conditions. Compare the wind speed, wave height, and visibility of each time period with the corresponding operation thresholds. Determine the time periods that simultaneously meet the wind speed threshold, wave height threshold, and visibility threshold as candidate operation time periods. Based on the risk characterization parameters, determine the priority of the targets to be maintained. Combined with the operation and maintenance resource parameters, match different maintenance objects with the preset maintenance operation types to the corresponding candidate operation time periods. Select an operation period and retain matching results that simultaneously meet the conditions of available vessels, personnel, equipment, and materials. Generate multiple executable maintenance plans that include operation period, maintenance object, and operation type. Based on risk characterization parameters, downtime impact parameters, and resource consumption parameters such as material consumption, personnel input, vessel occupancy, and downtime for each executable maintenance plan, determine the failure risk cost, resource consumption cost, and downtime impact cost for each plan, and obtain a comprehensive weighted evaluation value for the plan. Under the condition of satisfying structural reliability constraints, use heuristic optimization algorithms, reinforcement learning algorithms, or operations research solvers to determine the target maintenance plan that satisfies the structural reliability constraints and has the optimal evaluation value from multiple executable maintenance plans. The structural reliability constraints are used to limit the failure probability of the offshore wind power support structure after the maintenance operation to not exceed a preset safety threshold. Finally, output the target maintenance plan, including operation period, maintenance object, operation type, resource allocation results, estimated total operation cost, and expected risk reduction degree.

[0116] The first step in this embodiment is to construct a dynamic environmental corrosion model corresponding to the WT23 monopile based on the acquired environmental time-series data, and calculate the dynamic corrosion equivalent index and the real-time dynamic corrosion rate. First, the dynamic corrosion equivalent index is calculated using the formula: DCEI(t) = w1×S(t) + w2×T(t) + w3×RH(t) + w4×Cl(t) + C. In this formula, DCEI(t) is the dynamic corrosion equivalent index at time t, which is a core comprehensive indicator for quantifying the erosive ability of the target marine environment on the supporting structure. A higher value indicates stronger environmental corrosivity. In this embodiment, t is taken as the time corresponding to the 6th month of the 7th year of the monopile's service. S(t), T(t), RH(t), and Cl(t) are the normalized real-time values ​​of salinity, temperature, relative humidity, and chloride ion concentration, respectively, with a normalization interval of [0,1]. The original data are all from the collection results of environmental sensors. In this embodiment, the corresponding values ​​are 0.62, 0.58, 0.85, and 0.71, respectively; w1, w2, w3, and w4 are the weighting coefficients of the corresponding environmental factors, which are dimensionless parameters that satisfy w1+w2+w3+w4=1. The values ​​are obtained by calibration through historical corrosion data of the target sea area and material corrosion tests. In this embodiment, the corresponding values ​​are 0.2, 0.15, 0.23, and 0.42, respectively; C is a dimensionless material constant related to the substrate of the supporting structure (low-carbon steel for marine engineering), which is obtained by calibration through material electrochemical corrosion tests and reflects the intrinsic corrosion resistance properties of the substrate itself. In this embodiment, the value is 0.12; after calculation, the final calculated value of the dynamic corrosion equivalent index at time t in this embodiment is 1.28.

[0117] Based on the calculation results of the dynamic corrosion equivalent index, the real-time dynamic corrosion rate is calculated using the dynamic corrosion rate formula: Vcor(t) = BaseRate × DCEI(t) × CoatingFactor(t). In this formula, Vcor(t) is the real-time corrosion rate of the supporting structure substrate at time t, expressed in mm / year, reflecting the real-time development speed of structural corrosion damage. BaseRate is the baseline corrosion rate of the supporting structure substrate under standard marine conditions, expressed in mm / year. It is a fixed constant, calibrated through standard corrosion tests of marine engineering steel and historical corrosion data from the target sea area. In this embodiment... The value is 0.22 mm / year; DCEI(t) is the dynamic corrosion equivalent index calculated above, which is the environmental correction coefficient for the corrosion rate; CoatingFactor(t) is the degradation coefficient of the anti-corrosion coating of the supporting structure at time t, which is dimensionless and has a value range of [0,1]. A value of 1 represents that the coating is intact and has no degradation, and a value of 0 represents that the coating has completely failed. It is calculated by the coating damage identification results of the image acquisition device. In this embodiment, the coating is damaged in the tidal zone, and the corrosion is accelerated after the coating fails. The corresponding value is 1.13; According to the calculation, the dynamic corrosion rate at time t in this embodiment is 0.35 mm / year.

[0118] In the second step of this embodiment, the time-varying predicted corrosion depth of the WT23 monopile is calculated based on the calculated dynamic corrosion rate using the following formula: dpred(t) = ∫0tVcor(τ)dτ, where dpred(t) is the cumulative predicted corrosion depth of the supporting structure substrate from the time of structural commissioning to time t, in mm, directly reflecting the degree of corrosion damage to the structure; Vcor(τ) is the real-time corrosion rate of the structure at time τ, and the corrosion rate within the integration interval changes dynamically with environmental parameters and coating condition. In this embodiment, the average value within the integration interval is 0.35 mm / year; τ is the corrosion duration integral variable in years, and the integration interval is from the time of structural commissioning 0 to the calculation time t; t is the cumulative service time of the structure from commissioning to the calculation time in years, and in this embodiment, it is taken as 7 years; after calculation, the cumulative predicted corrosion depth at t=7 years in this embodiment is 2.45 mm.

[0119] In the third step of this embodiment, using structural response data as observational evidence, the parameters of the dynamic environmental corrosion model are calibrated online using the Bayesian update principle. The core formula is P(θ∣Z)=P(Z)P(Z∣θ)×P(θ), where P(θ∣Z) is the conditional probability distribution of the model parameter θ after calibration based on the monitoring data Z, which is the final output of the Bayesian update and represents the model parameter distribution that better reflects the actual state of the structure after calibration; θ is the set of core parameters of the corrosion model to be calibrated, including the weight coefficients w1-w4 in the dynamic corrosion equivalent index formula, the base corrosion rate BaseRate in the dynamic corrosion rate formula, and the coating degradation factor CoatingFactor(t); Z is the measured data of structural health monitoring, i.e., the observational evidence of the Bayesian update. In this embodiment, the measured value of the single pile vibration frequency and the measured corrosion depth of underwater detection are used as observational evidence; P (Z|θ) is the likelihood function, representing the probability of observed data Z occurring given a specific value of the model parameter θ. It reflects the degree of matching between the model prediction and the measured data. The higher the matching degree, the larger the likelihood function value. In this embodiment, the likelihood function is constructed using a normal distribution to match the deviation between the predicted corrosion depth and the measured corrosion depth. P(θ) is the prior probability distribution of the model parameters, representing the initial probability distribution of the model parameters θ before the introduction of new monitoring data. It is derived from the historical corrosion model parameter distribution of the same type of single pile in the same sea area. P(Z) is the normalization constant, which is the marginal probability of the observed data Z, used to ensure that the sum of the posterior probability distributions is 1. After Bayesian update calibration, this embodiment outputs the corrected probability distribution of the dynamic corrosion rate and the environmental factor weight coefficient. The deviation between the calibrated corrosion depth and the underwater measured value is reduced from 12.7% to 3.2%, significantly improving the prediction accuracy.

[0120] In the fourth step of this embodiment, based on the calibrated corrosion depth results, the cumulative damage degree of the WT23 single pile is calculated by coupling dynamic corrosion and fatigue damage effects. The core formula is D(t) = ∑dcritVcor(Δt) + ∑Nini, where D(t) is the total cumulative damage degree of the supporting structure at time t, dimensionless, with a value range of [0,1]. A value of 0 represents no damage to the structure, and a value of 1 represents the structure reaching the ultimate failure state; Vcor(Δt) is the average dynamic corrosion rate of the structure within the time step Δt, in mm / year. In this embodiment, Δt is taken as 1 year, corresponding to an average value of 0.35 mm / year; dcrit is the critical corrosion depth that leads to the ultimate failure state of the structure, in mm. Based on structural design parameters, material mechanical properties, and industry standards, this value is a fixed constant, and in this embodiment, it is taken as 6.8 mm; Δt is the unit time step for calculating corrosion damage; ni is the actual number of cycles under the i-th level alternating stress of the structure, dimensionless, obtained through statistical analysis of monitoring data from structural strain gauges and vibration sensors, and in this embodiment, the cumulative statistical value is 8.2 × 10⁶ cycles; Ni is the theoretical fatigue limit cycle number of the structural substrate under the i-th level alternating stress, dimensionless, determined through material SN curves and structural design standards, and is a fixed constant, and in this embodiment, it is taken as 2.3 × 10⁷ cycles at the corresponding stress level; Calculation shows that in this embodiment, at t = 7 years, the total cumulative damage of the WT23 single pile is 0.36.

[0121] In the fifth step of this embodiment, based on the calculated cumulative damage, the risk characterization parameters corresponding to the WT23 monopile are determined, including the remaining usable time, the probability of structural failure within the prediction period, and the structural risk level. First, the remaining usable time is calculated using the formula RUL=kd1-D(t), where RUL is the remaining service life of the supporting structure in years, representing the remaining service time from the current moment until the structure reaches its ultimate failure state; D(t) is the total cumulative damage at time t calculated earlier; and kd is the average annual damage growth rate within the future service period of the structure, in units of 1 / year, predicted using historical damage development data and calibrated dynamic corrosion rate, and in this embodiment, it is taken as 0.0781 / year. Calculations show that the remaining usable time of the WT23 monopile in this embodiment is 8.2 years.

[0122] Subsequently, the structural failure probability within the prediction period is calculated using the failure probability formula, Pf=P(D(t+T)≥1), where Pf is the probability that the supporting structure reaches the ultimate failure state within the prediction period T, with a value range of [0,1]; T is the time period for failure probability prediction, in years, determined by the wind farm operation and maintenance planning cycle, and in this embodiment, it is taken as 1 year; D(t+T) is the cumulative damage degree of the structure at the end of the prediction period T, and the probability distribution result is obtained through 1000 Monte Carlo simulations and damage evolution model calculations; according to the calculation, the structural failure probability of WT23 monopile within the 12-month prediction period in this embodiment is 6.8%.

[0123] In the claims, description, and accompanying drawings of this invention, the term "plural" refers to two or more. Unless otherwise explicitly defined, the terms "upper," "lower," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing the invention and simplifying the descriptive process, and are not intended to indicate or imply that the device or element referred to must have the described specific orientation, or be constructed and operated in a specific orientation. Therefore, these descriptions should not be construed as limiting the invention. The terms "connected," "installed," "fixed," etc., should be interpreted broadly. For example, "connected" can be a fixed connection between multiple objects, a detachable connection between multiple objects, or an integral connection; it can be a direct connection between multiple objects or an indirect connection between multiple objects through an intermediate medium. For those skilled in the art, the specific meaning of the above terms in this invention can be understood based on the specific circumstances described above.

[0124] In the claims, description, and accompanying drawings of this invention, the terms "one embodiment," "some embodiments," "specific embodiment," etc., refer to a specific feature, structure, material, or characteristic described in connection with that embodiment or example, which is included in at least one embodiment or example of the invention. In the claims, description, and accompanying drawings of this invention, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0125] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. 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.

Claims

1. A method for determining the health status of an offshore wind power support structure, characterized in that, include: Acquire multi-source heterogeneous data of the offshore wind power support structure, including environmental data, structural monitoring data, and structural foundation data; Spatiotemporal alignment and data fusion are performed on the multi-source heterogeneous data to obtain a multi-source synchronized dataset; Based on the environmental data in the multi-source synchronous dataset, the dynamic corrosion equivalent index of the offshore wind power support structure at each monitoring time is determined. Based on the basic corrosion rate in the structural foundation data, the dynamic corrosion equivalent index, and the coating degradation factor at the corresponding monitoring time determined by the structural monitoring data, the dynamic corrosion rate of the offshore wind power support structure at each monitoring time is determined. The time-varying corrosion depth is calculated based on the cumulative dynamic corrosion rate. The observations are determined based on the structural monitoring data. The observations are used as input to a Bayesian filter. The dynamic corrosion rate is calibrated based on the Bayesian filter to obtain the calibrated dynamic corrosion rate. The time-varying corrosion depth is updated based on the calibrated dynamic corrosion rate; Based on the updated time-varying corrosion depth and the stress cycle information characterized by the structural monitoring data, the corrosion and fatigue cumulative damage values ​​of the offshore wind power support structure are determined. Based on the correspondence between the cumulative corrosion and fatigue damage values ​​and the preset damage threshold range, the health status level of the offshore wind power support structure is determined and output.

2. The method for determining the health status of offshore wind power support structures according to claim 1, characterized in that, The environmental data includes time-series data on salinity, temperature, relative humidity, chloride ion concentration, and pH of the sea area where the offshore wind power support structure is located. The structural monitoring data includes vibration mode data, strain time history data, natural frequency data, and coating surface image data of the offshore wind power support structure. The structural basic data includes the design parameters, material properties, historical test records, and factory anti-corrosion scheme data of the offshore wind power support structure. Among them, the material properties include the basic corrosion rate under standard environment.

3. The method for determining the health status of offshore wind power support structures according to claim 1, characterized in that, The spatiotemporal alignment and data fusion of the multi-source heterogeneous data includes: Based on the sampling frequency and spatiotemporal reference of multi-source heterogeneous data, time axis alignment and spatial location matching are performed on different data sources; Based on the historical normal fluctuation range corresponding to the multi-source heterogeneous data, outliers in the data are removed, and missing values ​​are filled by linear interpolation or historical mean method. Based on the completed data volume and dimensions, the data is normalized to obtain a multi-source synchronous dataset with a unified spatiotemporal benchmark.

4. The method for determining the health status of offshore wind power support structures according to claim 1, characterized in that, The dynamic corrosion equivalent index is determined as follows: Based on the centralized normalized salinity, temperature, relative humidity, and chloride ion concentration values ​​of the multi-source synchronous dataset, and combined with the weighting coefficients of the corresponding environmental factors and the material constants of the steel in the structural basis data, the dynamic corrosion equivalent index at the corresponding monitoring time is determined by weighted fitting.

5. The method for determining the health status of offshore wind power support structures according to claim 4, characterized in that, The dynamic corrosion rate is determined as follows: Based on the basic corrosion rate of steel under standard environment, the dynamic corrosion equivalent index at the corresponding monitoring time, and the coating degradation factor at the corresponding monitoring time determined by the structural monitoring data, the dynamic corrosion rate of the offshore wind power support structure at the corresponding time is determined by product fitting.

6. The method for determining the health status of offshore wind power support structures according to claim 5, characterized in that, The coating degradation factor is determined as follows: From the coating surface image data in the structural monitoring data, at least one image feature is extracted from the crack length, blister area ratio, and rust area ratio. The degree of damage and degradation of the coating is determined based on the extracted image features, and then the coating degradation factor at the corresponding time is determined. When the coating is intact and undamaged, the coating degradation factor is set to the minimum preset value; when the coating is completely ineffective, the coating degradation factor is set to the maximum value of 1.

7. The method for determining the health status of offshore wind power support structures according to claim 1, characterized in that, The calibration of the dynamic corrosion rate based on Bayesian filtering includes: Based on the predicted damage state characterized by the dynamic corrosion rate and time-varying corrosion depth, a prior distribution of Bayesian filtering is established. Based on the changes in natural frequency, vibration modes, and strain response in the structural monitoring data, observations are constructed, and the likelihood functions corresponding to the observations are established. Based on the Bayesian filtering principle, and combining the prior distribution with the likelihood function, the posterior probability distribution of the true degradation state of the fitting structure is determined. The dynamic corrosion rate is corrected based on the posterior probability distribution, and the time-varying corrosion depth is updated based on the corrected dynamic corrosion rate.

8. A device for determining the health status of an offshore wind power support structure, characterized in that, include: The first acquisition module is used to acquire multi-source heterogeneous data of the offshore wind power support structure, including environmental data, structural monitoring data and structural foundation data. The data fusion module is used to perform spatiotemporal alignment and data fusion on the multi-source heterogeneous data to obtain a multi-source synchronized dataset; The corrosion index calculation module is used to determine the dynamic corrosion equivalent index of the offshore wind power support structure at each monitoring time based on the environmental data in the multi-source synchronous dataset. The corrosion rate determination module is used to determine the dynamic corrosion rate of the offshore wind power support structure at each monitoring time based on the basic corrosion rate in the structural basic data, the dynamic corrosion equivalent index, and the coating degradation factor at the corresponding monitoring time determined by the structural monitoring data. The corrosion depth calculation module is used to calculate the time-varying corrosion depth based on the dynamic corrosion rate. The Bayesian calibration module is used to determine the observations based on the structural monitoring data, use the observations as input to the Bayesian filter, and calibrate the dynamic corrosion rate based on the Bayesian filter to obtain the calibrated dynamic corrosion rate. A corrosion depth update module is used to update the time-varying corrosion depth according to the calibrated dynamic corrosion rate. The damage assessment module is used to determine the cumulative corrosion and fatigue damage value of the offshore wind power support structure based on the updated time-varying corrosion depth and the stress cycle information characterized by the structural monitoring data. The status output module is used to determine and output the health status level of the offshore wind power support structure based on the correspondence between the cumulative corrosion and fatigue damage values ​​and the preset damage threshold range.

9. A device for determining the health status of an offshore wind power support structure, characterized in that, include: processor; A memory containing programs or instructions, wherein the processor, when executing the programs or instructions in the memory, implements the steps of the method for determining the health status of offshore wind power support structures as described in any one of claims 1 to 7.

10. A readable storage medium, characterized in that, The readable storage medium stores a program or instructions that, when executed by a processor, implement the steps of the method for determining the health status of offshore wind power support structures as described in any one of claims 1 to 7.

Images