Method and device for predicting service life of structural member in natural exposure environment

Abstract

The invention discloses a method and a device for predicting the service life of a structural member in a natural exposure environment. The method comprises the following steps: carrying out information extraction on a coating sample to obtain appearance data, corrosion data and water permeability data of the coating sample; inputting the appearance data, the corrosion data and the water permeability data into simulation software, and performing corrosion simulation on the coating sample by utilizing the simulation software to obtain corrosion simulation data; and inputting the appearance data and the corrosion analog simulation data into a life prediction model, and processing the appearance data and the corrosion analog simulation data by using the life prediction model to obtain a life prediction result of the target structural member. The technical problems that in the related technology, in the atmosphere exposure environment, the structural part corrosion prediction experiment is high in dependency, large in resource consumption and long in period, and a high-corrosion-risk area cannot be accurately recognized through a traditional method are solved.

Description

Technical Field

[0001] This invention relates to the field of corrosion behavior prediction and protection technology, and more specifically, to a method and apparatus for predicting the lifespan of structural components under natural exposure environments. Background Technology

[0002] Currently, the assessment of corrosion behavior in structural components mainly relies on laboratory experiments and field exposure experiments. While these methods are intuitive and reliable, they have significant limitations:

[0003] 1) High cost: The experimental equipment and materials are expensive, especially for long-term atmospheric exposure experiments.

[0004] 2) Time-consuming: Obtaining corrosion data usually takes several years, which seriously affects the development speed of new coatings and protective measures.

[0005] 3) Difficulty in locating high-risk areas: Corrosion products may move over time, making it impossible to accurately identify which areas are most vulnerable to corrosion.

[0006] 4) Limitations of numerical simulation: Although numerical simulation (such as finite element analysis) can accelerate the prediction of corrosion behavior to some extent, its prediction accuracy is often not ideal due to the lack of sufficient consideration of the complexity of the actual corrosion process, especially in terms of coating details and environmental impact.

[0007] 5) Lack of application of data-driven technologies: Although data-based machine learning and deep learning technologies have shown great potential in many fields in recent years, their application in corrosion prediction and life assessment is still in its early stages. Existing methods have failed to fully utilize the advantages of these technologies, such as their ability to handle nonlinear relationships and learn from limited samples.

[0008] There is currently no effective solution to the above problems. Summary of the Invention

[0009] This invention provides a method and apparatus for predicting the lifespan of structural components under natural exposure environments, in order to at least solve the technical problems in related technologies where corrosion prediction experiments for structural components under atmospheric exposure environments are highly dependent on experiments, consume a lot of resources, have long cycles, and traditional methods cannot accurately identify high-corrosion-risk areas.

[0010] According to one aspect of the present invention, a method for predicting the lifespan of a structural component under natural exposure is provided, comprising: extracting information from a coating sample to obtain appearance data, corrosion data, and water permeability data of the coating sample, wherein the coating sample is selected from different coatings of a target structural component, and the target structural component is a structural component that has undergone aging for a predetermined period under natural exposure; inputting the appearance data, corrosion data, and water permeability data into simulation software to simulate corrosion of the coating sample using the simulation software to obtain corrosion simulation data; inputting the appearance data and corrosion simulation data into a lifespan prediction model to process the appearance data and corrosion simulation data using the lifespan prediction model to obtain a lifespan prediction result for the target structural component; wherein the lifespan prediction model is obtained by machine learning training using multiple sets of training data, each set of training data including: sample appearance data, sample corrosion simulation data, and a sample lifespan prediction result corresponding to the sample appearance data and the sample corrosion simulation data.

[0011] Optionally, information is extracted from the coating sample to obtain the appearance data of the coating sample, including at least one of the following: extracting the morphological information of the coating sample using an industrial camera; observing the surface roughness of the coating sample using a microscope; measuring the surface gloss of the coating sample using a gloss meter and measuring the thickness of the coating sample using a thickness gauge; and measuring the bonding force between the coating sample and the metal substrate using a pull-out test sample.

[0012] Optionally, information extraction is performed on the coating sample to obtain corrosion data of the coating sample, including: performing electrochemical impedance spectroscopy and polarization tests on the coating sample using a three-electrode system to obtain electrochemical test data; plotting an electrochemical impedance spectroscopy curve based on the electrochemical test data, and fitting the electrochemical impedance spectroscopy curve using an equivalent circuit to obtain the corrosion data, wherein the corrosion data includes: corrosion voltage, corrosion current density, coating resistance, and corrosion parameters.

[0013] Optionally, information extraction is performed on the coating sample to obtain the water permeability data of the coating sample, including: measuring the first weight of the coating sample under dry coating using a weighing method, and measuring the second weight of the coating sample under saturated coating using the weighing method; determining the water permeability data based on the first weight and the second weight.

[0014] Optionally, the appearance data, corrosion data, and water permeability data are input into simulation software to simulate corrosion of the coating sample using the simulation software, obtaining corrosion simulation data. This includes: generating a three-dimensional model of the coating sample using the simulation software based on the appearance data; determining the polarization relationship between current density and potential when the electrode material is working as the boundary condition of the electrode surface, and simultaneously setting the electrolyte conductivity; simulating the chlorine concentration change during water permeation in the simulation software based on the three-dimensional model, the boundary conditions, and the electrolyte conductivity, obtaining simulation results; generating a chlorine concentration change simulation graph based on the simulation results; and comparing the various chlorine concentration change simulation graphs to obtain the corrosion simulation data.

[0015] Optionally, the appearance data and the corrosion simulation data are input into a life prediction model to process the appearance data and the corrosion simulation data using the life prediction model to obtain the life prediction result of the target structural component. This includes: standardizing the appearance data and the corrosion simulation data using the life prediction model to correct or remove abnormal data in the appearance data and the corrosion simulation data to obtain standardized data; and processing the standardized data using the life prediction model to obtain the life prediction result.

[0016] Optionally, after inputting the appearance data and the corrosion simulation data into the life prediction model to process the appearance data and the corrosion simulation data using the life prediction model to obtain the life prediction result of the target structural component, the life prediction method further includes: determining the protection level of the coating corresponding to the coating sample for the target structural component based on the life prediction result.

[0017] According to another aspect of the present invention, a life prediction device for a structural component under natural exposure environment is also provided, comprising: an extraction unit for extracting information from a coating sample to obtain appearance data, corrosion data, and water permeability data of the coating sample, wherein the coating sample is selected from different coatings of a target structural component, and the target structural component is a structural component that has undergone aging for a predetermined period of time under natural exposure environment; a simulation unit for inputting the appearance data, corrosion data, and water permeability data into simulation software to simulate corrosion of the coating sample using the simulation software to obtain corrosion simulation data; and a processing unit for inputting the appearance data and corrosion simulation data into a life prediction model to process the appearance data and corrosion simulation data using the life prediction model to obtain a life prediction result for the target structural component; wherein the life prediction model is obtained by machine learning training using multiple sets of training data, each set of training data including: sample appearance data, sample corrosion simulation data, and sample life prediction result corresponding to the sample appearance data and sample corrosion simulation data.

[0018] Optionally, the extraction unit includes at least one of the following: an extraction module for extracting morphological information of the coating sample using an industrial camera; an observation module for observing the surface roughness of the coating sample using a microscope; a first measurement module for measuring the surface gloss of the coating sample using a gloss meter and measuring the thickness of the coating sample using a thickness gauge; and a second measurement module for measuring the bonding force between the coating sample and the metal substrate using a pull-out test specimen.

[0019] Optionally, the extraction unit includes: a testing module for performing electrochemical impedance spectroscopy and polarization tests on the coating sample using a three-electrode system to obtain electrochemical test data; and a fitting module for plotting an electrochemical impedance spectroscopy curve based on the electrochemical test data and fitting the electrochemical impedance spectroscopy curve using an equivalent circuit to obtain the corrosion data, wherein the corrosion data includes: corrosion voltage, corrosion current density, coating resistance, and corrosion parameters.

[0020] Optionally, the extraction unit includes: a third measurement module, used to measure the first weight of the coating sample under dry coating using a weighing method, and to measure the second weight of the coating sample under saturated coating using the weighing method; and a first determination module, used to determine the water permeability data based on the first weight and the second weight.

[0021] Optionally, the simulation unit includes: a first generation module, used to generate a three-dimensional model of the coating sample based on the appearance data using the simulation software; a second determination module, used to determine the polarization relationship between current density and potential when the electrode material is working as the boundary condition of the electrode surface, and simultaneously set the electrolyte conductivity; a simulation module, used to simulate the chlorine concentration change during the seepage process in the simulation software based on the three-dimensional model, the boundary condition, and the electrolyte conductivity, and obtain simulation results; a second generation module, used to generate a chlorine concentration change simulation diagram based on the simulation results; and a comparison module, used to compare the various chlorine concentration change simulation diagrams to obtain the corrosion simulation data.

[0022] Optionally, the processing unit includes: a data processing module, used to standardize the appearance data and the corrosion simulation data using the lifetime prediction model, so as to correct or remove abnormal data in the appearance data and the corrosion simulation data to obtain standardized data; and a processing module, used to process the standardized data through the lifetime prediction model to obtain the lifetime prediction result.

[0023] Optionally, the life prediction device further includes: a determination unit, configured to, after inputting the appearance data and the corrosion simulation data into the life prediction model, process the appearance data and the corrosion simulation data using the life prediction model to obtain the life prediction result of the target structural component, determine the protection level of the coating corresponding to the coating sample for the target structural component based on the life prediction result.

[0024] According to one aspect of the present invention, a structural component life prediction system is provided, the structural component life prediction system using the structural component life prediction method described above under natural exposure environment.

[0025] According to one aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium including a stored program, wherein the program executes the method for predicting the lifetime of a structure under natural exposure environment as described in any of the above embodiments.

[0026] According to one aspect of the present invention, a processor is provided for running a program, wherein the program executes the life prediction method for structural components under natural exposure environment as described in any of the above embodiments.

[0027] According to one aspect of the present invention, a computer program product is provided, including computer instructions that, when executed by a processor, perform the life prediction method for structural components under natural exposure environments as described in any one of the above embodiments.

[0028] In this embodiment of the invention, information is extracted from the coating sample to obtain its appearance data, corrosion data, and water permeability data. The coating sample is selected from different coatings on a target structural component, which is a structural component that has undergone aging for a predetermined period under natural exposure. The appearance data, corrosion data, and water permeability data are input into simulation software to simulate corrosion of the coating sample, obtaining corrosion simulation data. The appearance data and corrosion simulation data are then input into a lifetime prediction model to process them, obtaining a lifetime prediction result for the target structural component. The lifetime prediction model is trained using multiple sets of training data through machine learning. Each set of training data includes: sample appearance data, sample corrosion simulation data, and a corresponding lifetime prediction result. The above technical solution achieves the goal of rapidly and accurately predicting the corrosion behavior of structural components in complex atmospheric environments by combining numerical calculation and deep learning technologies. By utilizing the pattern recognition capabilities of deep learning models for the corrosion process, the most vulnerable parts of the structural components can be effectively identified. This significantly reduces the cost and time of assessing the corrosion behavior of structural components, improves the accuracy and reliability of predictions, reduces errors that may arise from simple numerical simulations, accurately defines high-risk corrosion areas, and clarifies the scope of corrosion-sensitive regions. Furthermore, it solves the technical problems in related technologies where corrosion prediction of structural components in atmospheric exposure environments is highly dependent on experiments, consumes a lot of resources, has a long cycle, and traditional methods cannot accurately identify high-risk corrosion areas. Attached Figure Description

[0029] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0030] Figure 1 This is a hardware structure block diagram of a mobile terminal for a method of predicting the lifetime of a structural component in a natural exposure environment according to an embodiment of the present invention.

[0031] Figure 2 This is a flowchart of a method for predicting the lifespan of a structural component under natural exposure conditions according to an embodiment of the present invention;

[0032] Figure 3 This is a schematic diagram of the Tafel curve of the surface of sample F2 according to an embodiment of the present invention;

[0033] Figure 4 This is a schematic diagram of the Tafel curve of the surface of sample F3 according to an embodiment of the present invention;

[0034] Figure 5 This is the Tafel curve of the surface of sample F4 according to an embodiment of the present invention;

[0035] Figure 6 These are Nyquist curves of the surfaces of samples F1 and F2 according to embodiments of the present invention;

[0036] Figure 7 This is a normality test plot of the average gloss quantile of F2 in the hyperparameter tuning according to an embodiment of the present invention;

[0037] Figure 8 The corrosion distribution cloud maps of the F1 and F2 samples after 8 years, predicted by machine learning according to an embodiment of the present invention.

[0038] Figure 9 These are laser confocal topography images of the surfaces of samples F1 and F3 according to an embodiment of the present invention;

[0039] Figure 10 This is a simulation diagram of the water permeability of the surface of samples F1 and F3 based on the finite element method according to an embodiment of the present invention.

[0040] Figure 11 This is the average gloss histogram of F3 during hyperparameter tuning according to an embodiment of the present invention;

[0041] Figure 12 These are gloss diagrams of the surfaces of samples F1 and F4 according to embodiments of the present invention;

[0042] Figure 13 These are Nyquist images of the surfaces of samples F1 and F4 according to embodiments of the present invention;

[0043] Figure 14 The corrosion distribution cloud maps of the F1 and F4 samples after 8 years, predicted by machine learning according to an embodiment of the present invention.

[0044] Figure 15 These are corrosion distribution cloud maps of the surfaces of samples F2, F3, and F4 10 years later, predicted by machine learning according to an embodiment of the present invention.

[0045] Figure 16 This is a schematic diagram of a life prediction device for a structural component under natural exposure environment according to an embodiment of the present invention.

[0046] The above figures include the following reference numerals:

[0047] 102. Processor; 104. Memory; 106. Transmission device; 108. Input / output device. Detailed Implementation

[0048] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0049] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0050] As described in the background section, in related technologies, corrosion prediction experiments for structural components under atmospheric exposure are highly dependent on, resource-intensive, and time-consuming, and traditional methods cannot accurately identify defects in high-corrosion-risk areas. This invention provides a method and apparatus for predicting the lifespan of structural components under natural exposure environments, as well as the structural component, a computer-readable storage medium, a processor, and a computer program product.

[0051] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0052] The methods and embodiments provided in this invention can be executed on a mobile terminal, a computer terminal, or a similar computing device. Taking running on a mobile terminal as an example, Figure 1 This is a hardware structure block diagram of a mobile terminal for a method of predicting the lifetime of structural components under natural exposure environment, according to an embodiment of the present invention. Figure 1 As shown, a mobile terminal may include one or more ( Figure 1 Only one is shown in the diagram. A processor 102 (which may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.) and a memory 104 for storing data are also shown. The mobile terminal may further include a transmission device 106 for communication functions and an input / output device 108. Those skilled in the art will understand that... Figure 1The structure shown is for illustrative purposes only and does not limit the structure of the mobile terminal described above. For example, the mobile terminal may also include components that are more... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0053] The memory 104 can be used to store computer programs, such as application software programs and modules, like the computer program corresponding to the lifespan prediction method for structural components under natural exposure environments in this embodiment of the invention. The processor 102 executes various functional applications and data processing by running the computer programs stored in the memory 104, thereby implementing the above-described method. The memory 104 may include high-speed random access memory and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to the mobile terminal via a network. Examples of the aforementioned networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof. The transmission device 106 is used to receive or send data via a network. Specific examples of the aforementioned networks may include wireless networks provided by the mobile terminal's communication provider. In one instance, the transmission device 106 includes a Network Interface Controller (NIC), which can be connected to other network devices via a base station to communicate with the Internet. In one example, the transmission device 106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0054] Example 1

[0055] According to an embodiment of the present invention, a method embodiment for predicting the lifespan of a structural component in a natural exposure environment is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0056] Figure 2 This is a flowchart of a method for predicting the lifespan of structural components under natural exposure environments according to an embodiment of the present invention, such as... Figure 2 As shown, the method includes the following steps:

[0057] Step S202: Extract information from the coating sample to obtain the appearance data, corrosion data and water permeability data of the coating sample. The coating sample is selected from different coatings of the target structural component. The target structural component is a structural component that has been aged for a predetermined time under natural exposure environment.

[0058] In this embodiment, F1 and F2 coating samples exposed to a marine atmospheric environment for four years were selected. Corrosion hotspots were observed on the samples to prepare for subsequent parameter measurements. The test samples mainly used metals with different coatings as electrodes, with sample dimensions of 250mm × 100mm × 5mm. The F1 coating consisted of primer + paint A, and the F2 coating consisted of primer + paint B. Coating appearance data were extracted: Surface morphology, surface roughness, corrosion depth, gloss, thickness, and adhesion were extracted using macroscopic characterization methods. Specifically, the macroscopic morphology of F1 and F2 samples was observed at 20x magnification; the surface roughness value Sa and corrosion depth were measured, and transverse and longitudinal corrosion depth curves were plotted. A handheld measuring device was used to measure the coating gloss and thickness. A 10mm diameter ingot was used to apply a 50N tensile force to the coating surface to test the coating adhesion. Electrochemical data extraction of coating samples: Electrochemical impedance spectroscopy (EIS) and polarization tests were performed on different coating samples using a three-electrode system, and the required corrosion data were extracted. Specifically, EIS and Tafel tests were performed on the samples at room temperature, with a test frequency range of 0.01 Hz to 100,000 Hz, a test voltage range of -0.2 V to +0.2 V, and a test electrode area of ​​1.28 cm². Nyquist and Tafel curves were plotted based on the test data (see...). Figure 3 , Figure 4 as well as Figure 5 , Figure 3 This is a schematic diagram of the Tafel curve of the sample F2 surface according to an embodiment of the present invention. The horizontal axis represents the corrosion current density i, and the vertical axis represents the corrosion potential E. Figure 4 This is a schematic diagram of the Tafel curve of the sample F3 surface according to an embodiment of the present invention, where the horizontal axis represents the corrosion current density i and the vertical axis represents the corrosion potential E. Figure 5 The Tafel curve of sample F4 according to an embodiment of the present invention (the horizontal axis represents the corrosion current density i, and the vertical axis represents the corrosion potential E) is obtained by fitting the curve through an equivalent circuit to calculate the corrosion voltage, corrosion current density, corrosion rate, coating resistance, and charge transfer resistance of the coated sample, which are the corrosion parameters required for simulation. Water permeability experimental parameter measurement: The water permeability parameters of the F1 and F2 coated samples required for corrosion simulation are measured.

[0059] Figure 6 These are Nyquist curves of the surfaces of samples F1 and F2 according to an embodiment of the present invention. The horizontal axis Z' represents the real impedance, and the vertical axis Z" represents the imaginary impedance, as shown below. Figure 6 As shown, (a) is the Nyquist curve of F1, and (b) is the Nyquist curve of F2. By comparing the Nyquist curves of different coating samples, their corrosion protection performance can be visually evaluated. For example, for the Nyquist curve of sample F1, the semicircle of the curve is small and the slope is gentle, indicating that the protective effect of the F1 coating is relatively weak, and corrosion reaction is more likely to occur. For the Nyquist curve of sample F2, the curve shows a large semicircle and a steep slope, which means that the F1 coating has a higher charge transfer resistance and coating resistance, and a stronger inhibitory effect on corrosion reaction. This information is crucial for understanding the corrosion protection mechanism of coatings, identifying easily corroded areas, and optimizing coating formulations.

[0060] The core of this method lies in quantifying the electrochemical properties of the coating samples. Electrochemical impedance spectroscopy (EIS) and polarization testing are commonly used techniques for evaluating the protective performance of coatings. EIS calculates the coating's impedance characteristics by applying a small-amplitude AC voltage to the coating surface and measuring the resulting current response. The magnitude of the impedance reflects the coating's ability to block corrosive media; the higher the impedance, the stronger the barrier performance. Polarization testing measures the change in current by altering the potential, thereby obtaining parameters such as corrosion current density and corrosion potential. These parameters reveal the coating's corrosion activity and stability at different potentials. Water permeability data is primarily measured using a gravimetric method, involving a weight comparison of the coating in both dry and wet states. The weight of a dry coating represents its initial state, while the weight of a water-saturated coating is its weight after fully absorbing water. The difference between the two and the rate of change reflect the coating's water absorption capacity and moisture transport characteristics.

[0061] By implementing this control method, the electrochemical impedance characteristics and water permeability of the coated samples can be obtained comprehensively and accurately, providing a solid data foundation for subsequent corrosion behavior prediction and lifetime assessment. Electrochemical data reveals the electrochemical interaction between the coating and the base metal, as well as the coating's barrier effect against corrosive media; while water permeability data supplements this by revealing the coating's characteristics in humid environments regarding the transport of moisture and corrosive media. The comprehensive analysis of these two types of data helps us to deeply understand the coating's protective mechanism, identify which coating materials or processing techniques have stronger resistance to specific corrosive environments, and thus provide a scientific basis for selecting the most suitable protective coating.

[0062] Step S204: Input the appearance data, corrosion data, and water penetration rate data into the simulation software to simulate the corrosion of the coating sample and obtain corrosion simulation data.

[0063] In this embodiment, all extracted data are imported into finite element software to plot seepage simulation curves and distribution cloud maps, and output the simulated corrosion data. Based on actual appearance measurements, three-dimensional models F1 and F2 are established at a 1:1 scale, with model substrate dimensions of 250mm × 100mm × 5mm, and coating thickness being the average of measured layer thicknesses. The models are meshed using free tetrahedral meshing, and the mesh size is set to extremely fine to improve convergence. The chloride ion concentration data measured electrochemically is imported into the seepage rate model in the finite element software, and the chloride ion concentration during the seepage process is analyzed. - Simulate the concentration change and plot Cl - Concentration change simulation graph; Cl concentration during water seepage of F1 and F2 coated samples - Concentration change curves were compared. Physical fields for atmospheric corrosion and water permeability were selected, and the required parameters were input. Atmospheric corrosion was modeled using a secondary current distribution, and the physical field interface was set to the secondary current distribution physical field. This physical field was used to solve for the potentials on the electrode domains of components F1 and F2. The water permeability was modeled using the diffusion coefficient. The governing equations for modeling were selected as transient with initialization, and the step size was set and calculations were performed. Parameters such as the exchange current density and corrosion rate of F1 and F2 were calculated using linear Butler-Volmer kinetic expressions.

[0064] In this method, simulation software is typically based on finite element analysis (FEA) technology, which can simulate the mechanical, thermal, fluid dynamics, and chemical reactions of complex systems through mathematical models and physical laws. The simulation software is used to simulate the corrosion process of coated samples under specific environmental conditions. By setting appropriate boundary conditions (such as the relationship between current density and potential in electrochemical reactions) and considering the influence of water penetration on corrosion, the software uses this data to establish a virtual corrosion environment to predict and simulate the corrosion behavior of the coated samples. Once all the necessary data is imported, the simulation software uses numerical calculation methods, such as the finite element method, to solve the partial differential equations in the corrosion process, thereby simulating the corrosion depth, corrosion product distribution, and chloride ion penetration over time.

[0065] By implementing this control method, the corrosion behavior of the coating samples is presented in a digital and visual form. Through corrosion simulation, the corrosion state of the samples at different time points can be obtained, including but not limited to the spatiotemporal distribution of corrosion depth and corrosion rate, as well as the penetration path and concentration gradient of chloride ions. This data not only helps to intuitively understand the degradation process of the coating in a corrosive environment, but also provides quantitative corrosion assessment indicators, which are crucial for guiding coating design optimization and evaluating its protective effectiveness.

[0066] Step S206: Input the appearance data and corrosion simulation data into the life prediction model to process the appearance data and corrosion simulation data using the life prediction model to obtain the life prediction result of the target structural component; wherein, the life prediction model is obtained by machine learning training using multiple sets of training data, and each set of training data includes: sample appearance data, sample corrosion simulation data, and sample life prediction result corresponding to the sample appearance data and sample corrosion simulation data.

[0067] In this embodiment, appearance data and simulation output data are imported into a machine learning model for accurate lifetime prediction, outputting a corrosion distribution cloud map. First, the extracted data is standardized, and outliers are corrected or removed using the IQR rule to ensure data integrity and consistency. XGBoost (…) is selected. The algorithm is coupled with linear regression as the core algorithm of the model, and hyperparameters are tuned through grid search (see [link]). Figure 7 , Figure 7 This is a normality test plot of the average gloss quantile of F2 in hyperparameter tuning according to an embodiment of the present invention, where the horizontal axis represents the theoretical quantile and the vertical axis represents the ordered value. The cross-entropy loss function is selected. This is used to measure the difference between the predicted probability distribution and the true distribution. Secondly, a corrosion lifetime prediction model based on machine learning, the remaining lifetime T model, is established: ,in, The critical corrosion depth of the material. Given the current corrosion depth, The corrosion rate is represented by the following model, based on the visual observation results from outdoor exposure experiments: , ,in, for Real-time appearance parameters For time series functions, , , , They are respectively Coating gloss at all times , , value. A corrosion rate exceeding 30% is considered a failure. Finally, the standardized data is input into the life prediction model for accurate life prediction. The output results are compared with the corrosion distribution cloud map to verify the model's accuracy. The accuracy of the machine learning prediction is further verified by comparing the output corrosion data with the electrochemical test data of samples at the current corrosion age. Finally, the predicted corrosion distribution cloud map is output to identify corrosion hotspots and optimize the protection method for different coatings. First, the corrosion results of the samples are observed, and the machine learning corrosion models of the surface corrosion of F1 and F2 coating samples after the same corrosion age are analyzed. Next, the corrosion parameters output by the corrosion model are observed, including corrosion potential, corrosion current density, and water penetration rate. The data from electrochemical tests and simulations are compared to observe the accuracy of the machine learning results. If incorrect, parameter optimization continues, using the cross-entropy loss function to reduce the difference between the probability distribution and the true distribution, thereby improving the model's accuracy. If correct, real-time coating data can be input for real-time life prediction, allowing for a more accurate comparison of the protective effects of F1 and F2 coatings.

[0068] Figure 8 This is a corrosion distribution cloud map of the F1 and F2 sample surfaces after 8 years, predicted by machine learning according to an embodiment of the present invention. The right axis represents the degree of corrosion; the darker the color, the more severe the corrosion. Figure 8 As shown, (a) is the corrosion distribution cloud map of sample F1 after 8 years, and (b) is the corrosion distribution cloud map of sample F2 after 8 years. Different colors or hues represent different corrosion depths or rates, thus revealing which areas are "hot spots" with more severe corrosion and which areas are relatively less affected by corrosion and maintain a better protective state. In this way, researchers and engineers can clearly identify which areas are high-risk corrosion zones and which areas are well protected, contributing to a deeper understanding of the corrosion protection performance of coatings.

[0069] This method builds its model based on a large historical training dataset, where each dataset contains specific appearance features, corrosion simulation data, and corresponding real service life results. The training process involves using advanced machine learning techniques such as XGBoost coupled with linear regression to optimize model parameters and minimize prediction errors. This allows the model to learn the complex relationship between the appearance properties, corrosion parameters, and service life of coating samples from the data. Appearance data reflects the physical state of the coating, including gloss, thickness, and surface roughness, while corrosion simulation data provides chemical dynamics information about corrosion rate, depth, and the corrosion environment. When these two types of data are input into the machine learning model, the model automatically seeks the intrinsic relationship between these data features and remaining service life. This relationship is often non-linear and influenced by multiple factors, including coating type, environmental conditions, and the properties of the corrosive medium. Through training, the model learns how to predict future corrosion trends and the remaining service time of structural components under these trends based on current appearance and corrosion data.

[0070] By implementing this control method, the predicted results are more closely aligned with reality, providing the remaining lifespan of structural components under specific conditions. This is of great significance for developing anti-corrosion maintenance plans and assessing asset value. The obtained lifespan predictions not only help engineers and decision-makers better understand the protective effectiveness of coatings but also guide them in making optimal maintenance decisions with limited resources. For example, when it is predicted that certain areas or specific types of coatings may face a shorter lifespan, inspections and repairs can be prioritized to avoid sudden failures that could lead to production disruptions or safety accidents.

[0071] As described above, in this embodiment, information is extracted from the coating sample to obtain its appearance data, corrosion data, and water permeability data. The coating sample is selected from different coatings on the target structural component, which is a structural component aged for a predetermined period under natural exposure. The appearance data, corrosion data, and water permeability data are input into simulation software to simulate corrosion of the coating sample, obtaining corrosion simulation data. The appearance data and corrosion simulation data are then input into a lifetime prediction model to process the appearance data and corrosion simulation data, obtaining the lifetime prediction result for the target structural component. The lifetime prediction model uses multiple sets of training data... Each set of training data obtained through machine learning training includes: sample appearance data, sample corrosion simulation data, and sample lifetime prediction results corresponding to the sample appearance data and sample corrosion simulation data. This achieves the goal of combining numerical calculation and deep learning technology to quickly and accurately predict the corrosion behavior of structural components in complex atmospheric environments. By utilizing the pattern recognition capability of the deep learning model for the corrosion process, it effectively identifies the parts of the structural components most susceptible to corrosion, thereby significantly reducing the cost and time of assessing the corrosion behavior of structural components, improving the accuracy and reliability of predictions, reducing the errors that may be generated by simple numerical simulation, accurately defining high-risk areas prone to corrosion, and clarifying the scope of corrosion-sensitive areas.

[0072] Therefore, the technical solution provided by the above embodiments of the present invention solves the technical problems in the related art where corrosion prediction experiments for structural components are highly dependent on atmospheric exposure, consume a lot of resources, have a long cycle, and traditional methods cannot accurately identify high corrosion risk areas.

[0073] According to the above embodiments of the present invention, information extraction is performed on the coating sample to obtain the appearance data of the coating sample, including at least one of the following: extracting the morphological information of the coating sample by an industrial camera; observing the surface roughness of the coating sample by a microscope; measuring the surface gloss of the coating sample by a gloss meter and measuring the thickness of the coating sample by a thickness gauge; and measuring the bonding force between the coating sample and the metal substrate by a pull-out test sample.

[0074] In this embodiment, F1 and F3 coating samples exposed to a marine atmospheric environment for four years were selected. Corrosion hotspots were observed on the samples to prepare for subsequent parameter measurements. The test samples mainly used metals with different coatings as electrodes, with sample dimensions of 250mm × 100mm × 5mm. The F1 coating consisted of primer + A coating, and the F3 coating consisted of primer + C coating. Macroscopic characterization methods were used to extract data on the coating's surface morphology, surface roughness, corrosion depth, gloss, thickness, and adhesion. Specifically, the macroscopic morphology of the F1 and F3 samples was observed at 20x magnification; the surface roughness value Sa and corrosion depth were measured, and transverse and longitudinal corrosion depth curves were plotted. A handheld measuring device was used to measure the coating's gloss and thickness. A 10mm diameter ingot was used to apply a 50N tensile force to the coating surface to test the coating's adhesion.

[0075] Figure 9 These are laser confocal topography images of the surfaces of samples F1 and F3 according to embodiments of the present invention, such as... Figure 9 As shown, (a) is a laser confocal microscopy (LCM) image of the surface of sample F1, and (b) is a LCM image of the surface of sample F2. The microscopic surface structures of the F1 and F2 coated samples after exposure to specific environmental conditions (such as a marine atmospheric environment) are presented. By comparing the images before and after exposure, the durability and degradation of the coatings can be evaluated. LCM images allow for a direct assessment of the performance of different coatings and the identification of coating defects and damage, which is crucial for coating selection and the formulation of anti-corrosion strategies.

[0076] In this method, an industrial camera acquires the macroscopic morphology of the coating through image analysis, such as cracks, blistering, and peeling. These features are intuitive indications of early coating failure. Microscopes, especially laser confocal microscopes, can observe the microstructure of the coating in depth, including surface roughness and localized corrosion depth. This information reflects the coating's microscopic defects and corrosion susceptibility. Gloss changes measured by a gloss meter are related to coating aging and corrosion levels, while a thickness gauge can directly detect changes in coating thickness. These two indicators together reflect the evolution of the coating's protective performance over time.

[0077] By implementing this control method, appearance data extraction not only provides a comprehensive assessment of the current state of the coated sample but also lays a solid data foundation for subsequent corrosion simulation and lifetime prediction. Quantifying the coating's morphology, gloss, thickness, and adhesion properties allows for accurate capture of changes in the coating's physical state, including its integrity, surface quality, and bonding with the metal substrate—all key indicators of its protective performance. Precise measurements of these data, combined with electrochemical test data, provide more accurate input for setting boundary conditions in finite element simulations, thereby improving the accuracy of corrosion simulation.

[0078] According to the above embodiments of the present invention, information extraction of the coating sample to obtain corrosion data of the coating sample includes: performing electrochemical impedance spectroscopy and polarization tests on the coating sample through a three-electrode system to obtain electrochemical test data; plotting an electrochemical impedance spectroscopy curve based on the electrochemical test data, and fitting the electrochemical impedance spectroscopy curve through an equivalent circuit to obtain corrosion data, wherein the corrosion data includes: corrosion voltage, corrosion current density, coating resistance, and corrosion parameters.

[0079] In this embodiment, EIS and Tafel tests were performed on the sample at room temperature, with a test frequency range of 0.01Hz-100000Hz, a test voltage range of -0.2V to +0.2V, and a test electrode area of ​​1.28cm². 2 Based on the test data, Nyquist and Tafel curves were plotted and fitted using an equivalent circuit. The corrosion voltage, corrosion current density, corrosion rate, coating resistance, and charge transfer resistance of the coated sample were calculated to obtain the corrosion parameters required for simulation.

[0080] In this method, corrosion data extraction from the coated samples is primarily performed using electrochemical testing techniques, specifically electrochemical impedance spectroscopy (EIS) and polarization curve testing in a three-electrode system. EIS is an advanced electrochemical testing method that measures the impedance response of the coating at different frequencies by applying a small-amplitude AC signal to the electrode surface, thereby obtaining the electrochemical characteristics of the coating-metal interface. Polarization testing involves changing the electrode potential and measuring the change in current density to obtain anodic and cathodic polarization curves, which are then used to analyze the kinetic characteristics of the corrosion process. The electrochemical data generated from these tests, such as EIS data, are further processed to plot Nyquist or Bode curves, which describe the changes in impedance, capacitance, and resistance of the electrochemical system with frequency. Then, the test curves are fitted using an equivalent circuit model to extract key corrosion parameters, including corrosion voltage, corrosion current density, and coating resistance. These parameters reflect the electrochemical stability of the coating and the corrosion tendency of the metal substrate.

[0081] By implementing this control method, the corrosion behavior of coated samples can be quantified, providing crucial parameters for machine learning lifetime prediction models. The impedance characteristics in the EIS test results are closely related to the coating's protective capability; low impedance often indicates higher corrosion activity. The corrosion current density in the polarization test is directly related to the corrosion rate, revealing the actual corrosion speed of the coating in a corrosive environment. The coating resistance reflects its ability to prevent the penetration of corrosive media and is a key indicator for assessing the long-term stability of the coating. Acquiring this corrosion data not only helps engineers understand the protective effect of coated samples under specific conditions but also allows for more accurate simulation and prediction of the coating's corrosion process by inputting it into finite element corrosion simulation models.

[0082] According to the above embodiments of the present invention, information extraction of the coating sample to obtain water permeability data of the coating sample includes: measuring the first weight of the coating sample under dry coating by weighing method, and measuring the second weight of the coating sample under saturated coating by weighing method; determining the water permeability data based on the first weight and the second weight.

[0083] In this embodiment, water permeability data for different types of coatings were extracted. The weights of the dry and water-saturated coatings were measured using a gravimetric method, and the saturated water content of the coatings was calculated. An electrochemical workstation was used to monitor the chloride ion concentration during the coating water permeation process in real time. The obtained data were then processed and plotted to map the chloride ion concentration at the monitoring points along the coating water permeation process. - Concentration change simulation curve.

[0084] This method is based on the physical process of water molecules penetrating through a coating into a metal substrate. First, under laboratory conditions, the weight of the coating sample in a dry state (dry coating weight) and in a fully water-saturated state (wet coating weight) are measured. The weight of the coating in the water-saturated state increases due to the absorption and adsorption of moisture. The principle of water penetration rate measurement lies in comparing these two sets of weight data. By calculating the difference between the dry and wet coating weights and combining this with the coating volume, the saturated water content of the coating can be determined.

[0085] By implementing this control method, the waterproof performance of the coating can be quantified, which is one of the important indicators for evaluating the corrosion resistance of the coating. High permeability means that the coating has a weak barrier effect against water and corrosive media, allowing corrosive media to easily penetrate the coating and reach the metal substrate, thus accelerating the corrosion process. Therefore, obtaining permeability data is crucial for identifying potential corrosion risks of the coating, optimizing coating design, and predicting the corrosion behavior of the coating in actual service environments. In finite element simulations, permeability data is used to set the boundary conditions of the corrosion physics interface, simulating the diffusion process of water molecules in the coating, providing more realistic environmental parameters for corrosion simulation, thereby improving the reliability of the simulation results.

[0086] According to the above embodiments of the present invention, appearance data, corrosion data, and water penetration rate data are input into simulation software to simulate corrosion of the coating sample and obtain corrosion simulation data. This includes: generating a three-dimensional model of the coating sample based on the appearance data using the simulation software; determining the polarization relationship between current density and potential during electrode material operation as the boundary condition of the electrode surface, and simultaneously setting the electrolyte conductivity; simulating the chlorine concentration change during water penetration in the simulation software based on the three-dimensional model, boundary conditions, and electrolyte conductivity, and obtaining simulation results; generating a chlorine element concentration change simulation graph based on the simulation results; and comparing the simulation graphs of each chlorine element concentration change to obtain corrosion simulation data.

[0087] In this embodiment, all extracted data are imported into finite element software to plot water seepage simulation curves and distribution cloud maps, and output the simulated corrosion data. Based on actual appearance measurements, three-dimensional models F1 and F3 are established at a 1:1 scale, with model substrate dimensions of 250mm × 100mm × 5mm, and coating thickness being the average of the measured layer thickness. The models are meshed using free tetrahedral meshing, and the mesh size is set to be extremely fine to improve convergence. Boundary conditions are set, using the polarization relationship between current density and potential during electrode material operation as the boundary conditions for the electrode surface, setting the electrolyte conductivity, and establishing piecewise linear and piecewise nonlinear models. Simultaneously, the chloride ion concentration data measured electrochemically is imported into the water seepage rate model in the finite element software, and the chloride ion concentration during the water seepage process is analyzed. - Simulate the concentration change and plot Cl - Concentration change simulation graph; Cl concentration during water seepage of F1 and F3 coated samples - Concentration change curves were compared. Physical fields for atmospheric corrosion and water permeability were selected, and the required parameters were input. Atmospheric corrosion was modeled using a secondary current distribution, and the physical field interface was set to the secondary current distribution physical field. This physical field was used to solve for the potentials on the electrode domains of components F1 and F3. The water permeability was modeled using the diffusion coefficient. The governing equations for modeling were selected as transient with initialization, and the step size was set and calculations were performed. Parameters such as the exchange current density and corrosion rate of F1 and F3 were calculated using linear Butler-Volmer kinetic expressions.

[0088] Figure 10 These are simulation diagrams of the water permeability of the surfaces of samples F1 and F3 based on finite element analysis according to an embodiment of the present invention, as shown below. Figure 10As shown, (a) is a simulated water penetration rate diagram of the F1 sample surface, and (b) is a simulated water penetration rate diagram of the F2 sample surface. The color depth or gradation visually reflects the degree of water penetration in different areas of the sample surface; darker colors indicate more severe water penetration, while lighter colors indicate effective suppression of water penetration. These predicted results, presented in image form, clearly reveal the water penetration distribution on the coated sample surface, providing crucial information for coating performance evaluation and optimization.

[0089] This method is based on mathematical modeling and numerical calculation of the electrochemical corrosion process. First, appearance data, such as coating thickness, morphology, and adhesion, are used to establish a three-dimensional model of the coated sample, and meshing is performed to simulate the physical properties of the metal and coating. Next, boundary conditions are set, including the polarization relationship between current density and potential of the electrode material under operating conditions, and the conductivity of the electrolyte; these parameters constitute the basic physical conditions of the corrosion environment. Then, using the physical field interface of the simulation software, particularly the atmospheric corrosion and permeability physical fields, corrosion parameters and permeability data are input to simulate the diffusion process of chloride ions in the coated sample, thereby calculating the potential distribution and chloride ion concentration changes of the sample under corrosive conditions. In this process, a secondary current distribution physical field is used to solve for the potential distribution on the electrode surface, while the diffusion coefficient is used to simulate chloride ion migration, ultimately generating a simulation diagram of chloride ion concentration changes. By comparing the simulation diagrams of different coated samples, corrosion simulation data can be obtained, including the potential and chloride ion concentration distribution in corrosion hotspot areas, as well as the dynamic trend of corrosion behavior.

[0090] By implementing this control method, the corrosion behavior of coated samples under specific environmental conditions can be effectively predicted and visualized, enabling precise location of corrosion hotspots. The corrosion simulation data not only includes simulation results of potential distribution and chloride ion concentration, but also reflects the coating's permeability, electrochemical activity, and the diffusion characteristics of the corrosive medium—all key indicators for evaluating coating protective effects and predicting corrosion progression. By comparing the simulation results of different coated samples, it is possible to visually identify which coating is more effective at preventing chloride ion penetration under specific conditions, and which coating provides a more stable potential in the metal substrate. This provides a quantitative reference for coating selection and corrosion protection design of structural components.

[0091] According to the above embodiments of the present invention, appearance data and corrosion simulation data are input into a life prediction model to process the appearance data and corrosion simulation data using the life prediction model to obtain the life prediction result of the target structural component. This includes: standardizing the appearance data and corrosion simulation data using the life prediction model to correct or remove abnormal data in the appearance data and corrosion simulation data to obtain standardized data; and processing the standardized data using the life prediction model to obtain the life prediction result.

[0092] In this embodiment, appearance data and simulation output data are imported into a machine learning model for accurate lifetime prediction, outputting a corrosion distribution cloud map. First, the extracted data is standardized, and outliers are corrected or removed using the IQR rule to ensure data integrity and consistency. XGBoost (…) is selected. The model uses a coupled algorithm of ) and linear regression as its core algorithm, and performs hyperparameter tuning through grid search, selecting the cross-entropy loss function ( This is used to measure the difference between the predicted probability distribution and the true distribution. Secondly, a corrosion lifetime prediction model based on machine learning, the remaining lifetime T model, is established: ,in, The critical corrosion depth of the material. Given the current corrosion depth, The corrosion rate is represented by the following model, based on the visual observation results from outdoor exposure experiments: , ,in, for Real-time appearance parameters For time series functions, , , , They are respectively Coating gloss at all times , , value. A corrosion rate exceeding 30% is considered a failure. Finally, the standardized data is input into the life prediction model for accurate life prediction. The output results are compared, with the current corrosion status of the structural component being compared to the corrosion distribution cloud map to verify the model's accuracy. The accuracy of the machine learning prediction is further verified by comparing the output corrosion data with the electrochemical test data of samples at the current corrosion age. Finally, the predicted corrosion distribution cloud map is output to identify corrosion hotspots and optimize the protection method for different coatings. First, the corrosion results of the samples are observed, and the machine learning corrosion models of the surface corrosion of F1 and F3 coating samples after the same corrosion age are analyzed. Next, the corrosion parameters output by the corrosion model are observed, including corrosion potential, corrosion current density, and water penetration rate. The data from electrochemical tests and simulations are compared to observe the accuracy of the machine learning results. If incorrect, parameter optimization continues, using the cross-entropy loss function to reduce the difference between the probability distribution and the true distribution, thereby improving the model's accuracy. If correct, real-time coating data can be input for real-time life prediction, allowing for a more accurate comparison of the protective effects of F1 and F3 coatings.

[0093] Figure 11 This is a histogram of the average gloss of F3 in the hyperparameter tuning according to an embodiment of the present invention. The horizontal axis represents the average gloss, and the vertical axis represents the number of samples with gloss equal to the average gloss. Figure 11 As shown, the histogram groups the gloss data according to certain intervals, and the frequency of occurrence of each group is displayed in the form of a bar chart. By observing the height of these bars, one can understand the distribution of gloss values ​​in the dataset. Figure 11 The histogram shows particularly high gloss levels within a certain range, indicating that the gloss of the coating on the F3 sample surface is concentrated within this range, reflecting a general trend in coating variation during corrosion. Histograms help identify outliers and patterns in the data; this information allows for hyperparameter tuning, improving the predictive performance of machine learning models.

[0094] This method utilizes finite element simulation technology, combining the appearance values, electrochemical characteristics, and water permeability data of the coated samples to construct a corrosion model under a virtual environment. First, a three-dimensional model of the coated sample is established based on the appearance data, ensuring that the model accurately reflects the physical properties of the coating, including thickness, morphology, and adhesion. Next, by setting boundary conditions, such as the potential-current density polarization relationship of the electrode material and the conductivity of the electrolyte, the electrochemical response of the coating under atmospheric corrosion conditions is simulated. Specifically, the permeability of the coating to the corrosive medium is explored through the simulation of water permeability data and chloride ion concentration, which is one of the key factors in evaluating the coating's protective effect. In the selection of the physical field interface, a secondary current distribution method is used for atmospheric corrosion to more accurately describe the potential distribution, while the water permeability is simulated using the diffusion coefficient to simulate the migration of chloride ions in the coating. These two methods work together in the corrosion model to generate a dynamic simulation diagram of the potential distribution and chloride ion concentration changes.

[0095] By implementing this control method, the evaluation level of the corrosion protection performance of coated samples under specific environmental conditions can be significantly improved. Through visualizing corrosion simulation results, such as potential distribution maps and chloride ion concentration change maps, researchers can intuitively identify which type of coating is more effective at blocking chloride ion penetration, and under which coating the potential of the metal substrate is more stable—two important criteria for measuring the protective effect of the coating. This quantitative analysis method greatly facilitates the selection and optimization of coating materials and is of great significance for the corrosion protection design of structural components. It can not only predict the long-term performance of the coating but also anticipate potential corrosion hotspots in structural components, thus providing strategic guidance for corrosion prevention and control.

[0096] According to the above embodiments of the present invention, after inputting appearance data and corrosion simulation data into the life prediction model, and processing the appearance data and corrosion simulation data using the life prediction model to obtain the life prediction result of the target structural component, the life prediction method further includes: determining the protection level of the coating corresponding to the coating sample for the target structural component based on the life prediction result.

[0097] In this embodiment, a quantitative assessment of corrosion behavior and a comprehensive evaluation of material properties are combined.

[0098] The determination of the protection level using this method is based on two considerations: first, the predicted corrosion life of the coating sample under specific environmental conditions, i.e., the length of time the coating can effectively protect the metal substrate from corrosion; and second, the changes in appearance and the degree of performance degradation of the coating sample during the corrosion process. By comparing the evolution of corrosion degree and appearance parameters of different coating samples after the same exposure years, the protective effect of the coating is quantified. This process is essentially a comprehensive evaluation of the corrosion resistance of the coating sample, transforming the life prediction results into an intuitive protection level, which facilitates engineers and decision-makers in understanding and selecting the most suitable anti-corrosion coating.

[0099] By implementing this control method, the lifespan prediction results are converted into protection levels, making the evaluation of coating corrosion resistance more standardized and easier to understand. This facilitates the rapid selection of coating types suitable for specific environments and usage requirements. Secondly, determining the protection level provides clear guidance for the corrosion protection design of structural components. Engineers can select coatings with appropriate protection levels based on the expected service life and required safety margins, thereby optimizing the design and reducing costs. Finally, this method enhances the scientific rigor and objectivity of coating selection, avoiding misjudgments that may result from relying solely on experience or preliminary experimental results, and promoting the precise application and continuous innovation of coating technology in the field of corrosion protection for metal structural components.

[0100] As described above, the technical solution provided by the embodiments of the present invention, compared with traditional methods that mainly rely on physical experiments, integrates corrosion numerical simulation and machine learning techniques to construct a data-driven prediction model. This enables efficient and accurate evaluation of the corrosion resistance of different coatings in a short time and achieves reliable location of corrosion-prone areas. This not only significantly reduces R&D and testing costs but also significantly improves the accuracy of corrosion behavior prediction by correcting numerical simulation biases with intelligent algorithms. This provides quantitative and efficient technical support for the corrosion protection design, maintenance strategy formulation, and service life assessment of structural components.

[0101] To increase the diversity of the embodiments, F1 and F4 coating samples exposed to the marine atmospheric environment for four years were selected, and corrosion hot spots were observed on the samples. The sample size was 250mm×100mm×5mm, where F1 coating was primer + A coating and F4 coating was primer + D coating. Figure 12 These are gloss maps of the surfaces of samples F1 and F4 according to embodiments of the present invention, such as... Figure 12 As shown, (a) is the gloss map of sample F1, and (b) is the gloss map of sample F2. The changes in surface gloss of the F1 and F2 coated samples from their initial state to after a certain period of environmental exposure are illustrated (F1 has a higher gloss than F2). A decrease in gloss usually indicates wear, corrosion, or other forms of degradation on the coating surface, which directly affects the coating's ability to protect the metal substrate. By comparing the gloss maps of the two samples, the differences in resistance to environmental erosion between the two coatings can be clearly seen.

[0102] Figure 13 These are Nyquist curves of the surfaces of samples F1 and F4 according to embodiments of the present invention. The horizontal axis Z' represents the real impedance, and the vertical axis Z" represents the imaginary impedance, as shown below. Figure 13 As shown, (a) is the Nyquist curve for F1, and (b) is the Nyquist curve for F4. For the Nyquist curve of sample F1, the smaller semicircle and gentler slope indicate that the protective effect of the F1 coating is relatively weak, and corrosion is more likely to occur. For the Nyquist curve of sample F4, the larger semicircle and steeper slope suggest that the F1 coating has higher charge transfer resistance and coating resistance, resulting in stronger inhibition of corrosion. This information is crucial for understanding the corrosion protection mechanism of coatings, identifying easily corroded areas, and optimizing coating formulations.

[0103] Figure 14 This is a corrosion distribution cloud map of the F1 and F4 sample surfaces after 8 years, predicted by machine learning according to an embodiment of the present invention. The right axis represents the degree of corrosion; the darker the color, the more severe the corrosion. Figure 14 As shown, (a) is the corrosion distribution cloud map of the F1 sample surface after 8 years, and (b) is the corrosion distribution cloud map of the F4 sample surface after 8 years. Different colors or hues represent different corrosion depths or rates, thus revealing which areas are "hot spots" with more severe corrosion and which areas are relatively less affected by corrosion and maintain a better protective state. In this way, researchers and engineers can clearly identify which areas are high-corrosion-risk zones and which areas are well protected, contributing to a deeper understanding of the corrosion protection performance of the coating.

[0104] Figure 15This is a corrosion distribution cloud map of the surfaces of samples F2, F3, and F4 10 years later, predicted by machine learning according to an embodiment of the present invention. The right axis represents the degree of corrosion; the darker the color, the more severe the corrosion. Figure 15 As shown, (a), (b), and (c) are corrosion distribution cloud maps of samples F2, F3, and F4 after 10 years, respectively. Similarly, the depth of color represents the difference in corrosion degree; generally, darker areas indicate more severe corrosion, while lighter areas indicate less severe corrosion or effective protection. It is evident that F3 exhibits the most severe corrosion. By comparing the corrosion cloud maps of different samples, the long-term protective effectiveness of various coatings can be visually compared, identifying which coating is more suitable for use in a specific environment.

[0105] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0106] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods of the various embodiments of this application.

[0107] Example 2

[0108] According to embodiments of the present invention, an apparatus for predicting the lifespan of structural components under natural exposure environments, used for implementing the above-described method for predicting the lifespan of structural components under natural exposure environments, is also provided. Figure 16 This is a schematic diagram of a life prediction device for a structural component under natural exposure environment according to an embodiment of the present invention, as shown below. Figure 16 As shown, the device includes an extraction unit 1601, a simulation unit 1603, and a processing unit 1605. The device will now be described in detail.

[0109] Extraction unit 1601 is used to extract information from coating samples to obtain appearance data, corrosion data and water permeability data of coating samples. The coating samples are selected from different coatings of the target structural component, which is a structural component that has been aged for a predetermined time under natural exposure environment.

[0110] The simulation unit 1603 is used to input appearance data, corrosion data and water penetration rate data into the simulation software, so as to use the simulation software to simulate the corrosion of the coating sample and obtain corrosion simulation data.

[0111] The processing unit 1605 is used to input appearance data and corrosion simulation data into the life prediction model, so as to process the appearance data and corrosion simulation data using the life prediction model to obtain the life prediction result of the target structural component; wherein, the life prediction model is trained by machine learning using multiple sets of training data, and each set of training data includes: sample appearance data, sample corrosion simulation data, and sample life prediction results corresponding to the sample appearance data and sample corrosion simulation data.

[0112] It should be noted that the extraction unit 1601, simulation unit 1603 and processing unit 1605 mentioned above correspond to steps S202 to S206 in the above embodiments. The three units and the corresponding steps implement the same instances and application scenarios, but are not limited to the content disclosed in the above embodiments.

[0113] As can be seen from the above, in the scheme described in the above embodiments of the present invention, an extraction unit can be used to extract information from the coating sample to obtain the appearance data, corrosion data, and water permeability data of the coating sample. The coating sample is selected from different coatings on the target structural component, which is a structural component that has undergone aging for a predetermined period under natural exposure. A simulation unit is used to input the appearance data, corrosion data, and water permeability data into simulation software to simulate corrosion of the coating sample and obtain corrosion simulation data. A processing unit is used to input the appearance data and corrosion simulation data into a lifetime prediction model to process the appearance data and corrosion simulation data to obtain the lifetime prediction result of the target structural component. The lifetime prediction model is obtained by machine learning training using multiple sets of training data. Each set of training data includes: sample appearance data, sample corrosion simulation data, and the sample lifetime prediction result corresponding to the sample appearance data and sample corrosion simulation data. The above scheme achieves the goal of rapidly and accurately predicting the corrosion behavior of structural components in complex atmospheric environments by combining numerical calculation and deep learning technologies. By utilizing the pattern recognition capabilities of deep learning models for the corrosion process, it effectively identifies the most susceptible parts of the structural components to corrosion, thereby significantly reducing the cost and time of assessing the corrosion behavior of structural components, improving the accuracy and reliability of predictions, reducing the errors that may be generated by simple numerical simulations, accurately defining high-risk corrosion areas, and clarifying the scope of corrosion-sensitive areas. This solves the technical problems in related technologies where corrosion prediction of structural components in atmospheric exposure environments is highly dependent on experiments, consumes a lot of resources, has a long cycle, and traditional methods cannot accurately identify high-corrosion-risk areas.

[0114] Optionally, the extraction unit includes at least one of the following: an extraction module for extracting morphological information of the coating sample using an industrial camera; an observation module for observing the surface roughness of the coating sample using a microscope; a first measurement module for measuring the surface gloss of the coating sample using a gloss meter and measuring the thickness of the coating sample using a thickness gauge; and a second measurement module for measuring the bonding force between the coating sample and the metal substrate using a pull-out test specimen.

[0115] Optionally, the extraction unit includes: a testing module for performing electrochemical impedance spectroscopy and polarization tests on the coating sample using a three-electrode system to obtain electrochemical test data; and a fitting module for plotting an electrochemical impedance spectroscopy curve based on the electrochemical test data and fitting the electrochemical impedance spectroscopy curve using an equivalent circuit to obtain corrosion data, wherein the corrosion data includes: corrosion voltage, corrosion current density, coating resistance, and corrosion parameters.

[0116] Optionally, the extraction unit includes: a third measurement module for measuring the first weight of the coating sample under dry coating by weighing method, and measuring the second weight of the coating sample under saturated coating by weighing method; and a first determination module for determining water permeability data based on the first weight and the second weight.

[0117] Optionally, the simulation unit includes: a first generation module for generating a three-dimensional model of the coating sample based on appearance data using simulation software; a second determination module for determining the polarization relationship between current density and potential when the electrode material is working as the boundary condition of the electrode surface, and setting the electrolyte conductivity; a simulation module for simulating the chlorine concentration change during water seepage in the simulation software based on the three-dimensional model, boundary conditions, and electrolyte conductivity, and obtaining simulation results; a second generation module for generating a chlorine element concentration change simulation diagram based on the simulation results; and a comparison module for comparing the chlorine element concentration change simulation diagrams to obtain corrosion simulation data.

[0118] Optionally, the processing unit includes: a data processing module, used to standardize the appearance data and corrosion simulation data using a lifetime prediction model, so as to correct or remove abnormal data in the appearance data and corrosion simulation data to obtain standardized data; and a processing module, used to process the standardized data through the lifetime prediction model to obtain lifetime prediction results.

[0119] Optionally, the life prediction device further includes: a determination unit, used to determine the protection level of the coating sample for the target structure after inputting appearance data and corrosion simulation data into the life prediction model, processing the appearance data and corrosion simulation data using the life prediction model to obtain the life prediction result of the target structure, and then determining the protection level of the coating sample for the target structure based on the life prediction result.

[0120] According to one aspect of the present invention, a structural component life prediction system is provided, which uses the structural component life prediction method described above for natural exposure environment.

[0121] According to one aspect of the present invention, a processor is provided for running a program, wherein the program executes the life prediction method for any of the above-described structures under natural exposure environment.

[0122] According to one aspect of the present invention, a computer program product is provided, including computer instructions, which, when executed by a processor, perform a method for predicting the lifetime of a structure under natural exposure conditions, as described above.

[0123] According to one aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium including a stored program, wherein the program executes the method for predicting the lifetime of a structure under natural exposure environment as described above.

[0124] In the above embodiments of the present invention, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0125] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual couplings, direct couplings, or communication connections may be through some interfaces; indirect couplings or communication connections between units or modules may be electrical or other forms.

[0126] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0127] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0128] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0129] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0130] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for predicting the lifespan of a structural component under natural exposure conditions, characterized in that, include: Information is extracted from the coating sample to obtain the appearance data, corrosion data and water permeability data of the coating sample. The coating sample is selected from different coatings of the target structural component, which is a structural component that has been aged for a predetermined time under natural exposure. The appearance data, corrosion data, and water permeability data are input into the simulation software to simulate the corrosion of the coating sample and obtain corrosion simulation data. The appearance data and the corrosion simulation data are input into the life prediction model to process the appearance data and the corrosion simulation data using the life prediction model, thereby obtaining the life prediction result of the target structural component. The life prediction model is trained using multiple sets of training data through machine learning. Each set of training data includes: sample appearance data, sample corrosion simulation data, and sample life prediction results corresponding to the sample appearance data and the sample corrosion simulation data.

2. The method for predicting the lifespan of structural components under natural exposure environments according to claim 1, characterized in that, Information is extracted from the coating sample to obtain the appearance data of the coating sample, including at least one of the following: The morphological information of the coated sample was extracted using an industrial camera; The surface roughness of the coated sample was observed using a microscope. The surface gloss of the coated sample was measured using a gloss meter, and the thickness of the coated sample was measured using a thickness gauge. The adhesion between the coated sample and the metal substrate was measured by pull-out test.

3. The method for predicting the lifespan of structural components under natural exposure environments according to claim 1, characterized in that, Information is extracted from the coating sample to obtain corrosion data of the coating sample, including: Electrochemical impedance spectroscopy and polarization tests were performed on the coated sample using a three-electrode system to obtain electrochemical test data. Electrochemical impedance spectroscopy (EIS) curves are plotted based on the electrochemical test data, and the EIS curves are fitted using an equivalent circuit to obtain the corrosion data, which includes corrosion voltage, corrosion current density, coating resistance, and corrosion parameters.

4. The method for predicting the lifespan of structural components under natural exposure environments according to claim 1, characterized in that, Information is extracted from the coating sample to obtain the water permeability data of the coating sample, including: The first weight of the coating sample under dry coating is measured using a weighing method, and the second weight of the coating sample under saturated coating is measured using the same weighing method. The permeability data is determined based on the first weight and the second weight.

5. The method for predicting the lifespan of structural components under natural exposure environments according to claim 1, characterized in that, The appearance data, corrosion data, and water permeability data are input into simulation software to simulate the corrosion of the coating sample, obtaining corrosion simulation data, including: The simulation software is used to generate a three-dimensional model of the coating sample based on the appearance data; The polarization relationship between current density and potential when the electrode material is working is determined as the boundary condition of the electrode surface, and the electrolyte conductivity is set at the same time. Based on the three-dimensional model, the boundary conditions, and the electrolyte conductivity, the change in chlorine concentration during the seepage process was simulated in the simulation software, and the simulation results were obtained. Based on the simulation results, a simulation graph of the chloride element concentration change is generated; The corrosion simulation data are obtained by comparing the simulation graphs of the concentration changes of each chlorine element.

6. The method for predicting the lifespan of structural components under natural exposure environments according to claim 1, characterized in that, The appearance data and corrosion simulation data are input into the life prediction model to process the appearance data and corrosion simulation data using the life prediction model, thereby obtaining the life prediction result of the target structural component, including: The appearance data and corrosion simulation data are standardized using the life prediction model to correct or remove abnormal data in the appearance data and corrosion simulation data, thereby obtaining standardized data. The standardized data is processed by the lifespan prediction model to obtain the lifespan prediction result.

7. The method for predicting the lifespan of structural components under natural exposure environments according to claim 1, characterized in that, After inputting the appearance data and the corrosion simulation data into the life prediction model, and processing the appearance data and the corrosion simulation data using the life prediction model to obtain the life prediction result of the target structural component, the life prediction method further includes: The protection level of the coating on the target structure corresponding to the coating sample is determined based on the life prediction results.

8. A device for predicting the lifespan of a structural component under natural exposure conditions, characterized in that, include: The extraction unit is used to extract information from the coating sample to obtain the appearance data, corrosion data and water permeability data of the coating sample. The coating sample is selected from different coatings of the target structural component, which is a structural component that has been aged for a predetermined time under natural exposure. The simulation unit is used to input the appearance data, the corrosion data and the water penetration rate data into the simulation software, so as to use the simulation software to simulate the corrosion of the coating sample and obtain corrosion simulation data. The processing unit is configured to input the appearance data and the corrosion simulation data into the life prediction model, so as to process the appearance data and the corrosion simulation data using the life prediction model to obtain the life prediction result of the target structural component; wherein, the life prediction model is obtained by machine learning training using multiple sets of training data, each of the multiple sets of training data including: sample appearance data, sample corrosion simulation data, and sample life prediction result corresponding to the sample appearance data and the sample corrosion simulation data.

9. A structural component life prediction system, characterized in that, The structural component life prediction system uses the life prediction method for structural components under natural exposure environment as described in any one of claims 1 to 7.

10. A computer program product comprising computer instructions, characterized in that, When the computer instructions are executed by the processor, they perform the life prediction method for structural components under natural exposure environments as described in any one of claims 1 to 7.

Images