Steel structure anticorrosion evaluation system based on environmental data and AI

Abstract

The invention relates to the technical field of structure health monitoring and artificial intelligence crossing, and discloses a steel structure anticorrosion evaluation system based on environmental data and AI. The system comprises a data sensing module, an environment index generation module, an impact function generation module, a stress-corrosion coupling effect model construction module, a steel structure service life prediction module, a grading early warning module and a system self-learning module. According to the invention, corrosion impact of environmental emergencies can be quantified, and advanced early warning is realized; real-time stress monitoring data are directly incorporated into a corrosion life evaluation model for the first time, and the stress corrosion coupling effect is accurately described; modeling and prediction are carried out for the unique environment and stress state of each specific structure, and the result has more guiding significance. According to the invention, the transformation from planned maintenance to condition-based maintenance is realized, the waste caused by excessive maintenance is avoided, the risk caused by insufficient maintenance can be prevented, and the whole life cycle cost is obviously reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of structural health monitoring and artificial intelligence, and more particularly to a steel structure corrosion prevention evaluation system based on environmental data and AI. BACKGROUND

[0002] Steel structures are widely used in the fields of bridges, ports, wind power towers, chemical equipment, etc., but their corrosion problems are the core difficulties affecting the safety and service life of the structures.

[0003] Traditional corrosion prevention evaluation methods mainly rely on periodic manual inspection and local non-destructive testing. The conventional approach includes sampling evaluation of the corrosion state of the structure through visual inspection, coating thickness gauge, corrosion coupon, ultrasonic flaw detection, etc. At the same time, some advanced systems have introduced environmental sensors such as temperature and humidity, salt fog concentration monitoring to collect corrosion environment data, and established the correlation between corrosion rate and environmental factors based on statistical methods or empirical models, thereby realizing preliminary prediction of the corrosion trend and life estimation of the structure.

[0004] The existing evaluation methods also have the following defects: The existing technology lacks effective identification and quantification of environmental mutation events. Its evaluation model is mostly based on long-term average environmental data, which cannot capture and quantify the impact of short-term severe environmental changes such as heavy rain, salt fog attack, and chemical leakage on the corrosion process, resulting in a lack of sensitivity to accelerated corrosion failure risks caused by sudden events, delayed warning, and often only discovered through regular inspections after damage occurs. The existing life prediction methods generally fail to incorporate real-time stress monitoring data directly into the evaluation model, usually using conservative and fixed safety factors to roughly estimate the impact of stress, or completely ignoring the coupling effect of stress and environment, which results in an inability to accurately depict the actual risks of stress corrosion cracking and corrosion fatigue, making the life prediction results significantly biased, overly conservative or overly optimistic, and unable to provide accurate basis for decision-making. The existing evaluation system highly depends on industry-wide standards, specifications, or average empirical models based on large amounts of data statistics, which fail to fully consider the microenvironmental differences of specific structures, the uniqueness of load history, and the individualized characteristics of material state, leading to generalized evaluation results that cannot accurately reflect the true state of specific objects and have limited guidance significance. The existing maintenance strategy mainly adopts a planned maintenance mode based on fixed cycles, which can easily produce two drawbacks: one is over-maintenance, which unnecessarily repairs and replaces well-conditioned components, wasting human and material resources; the other is insufficient maintenance, which fails to timely detect hidden damage and miss the best maintenance window, ultimately leading to small problems becoming big problems and significantly increasing repair costs and safety risks.

[0005] Therefore, a method with the characteristics of foreknowledge, accurate estimation, individualized difference consideration and cost reduction and benefit increase is needed to solve the above problems. SUMMARY

[0006] In order to overcome the above defects of the prior art, the present application provides a steel structure corrosion prevention evaluation system based on environmental data and AI to solve the problems existing in the background art.

[0007] To achieve the above object, the present application provides the following technical scheme: a steel structure corrosion prevention evaluation system based on environmental data and AI, comprising: A data perception module is used to collect real-time environmental data and steel structure state data of the steel structure; An environmental index generation module is used to generate a real-time environmental corrosivity index through weighted calculation according to the current environmental data of the steel structure; An impact function generation module is used to identify environmental mutations, combine the environmental corrosivity index change intensity and duration, and construct an environmental mutation impact function to calculate the additional corrosion amount; A stress-corrosion coupling effect model construction module is used to select an algorithm and train a stress-corrosion coupling effect model according to the real-time state data of the steel structure, the real-time environmental corrosivity index and the material properties, and combine the impact function to output the actual corrosion rate; A steel structure life prediction module is used to calculate the current corrosion cumulative damage of the steel structure according to the actual corrosion rate and perform life prediction; A graded early warning module is used to combine the output results of each model and sensor data to perform graded early warning and provide visual decision support.

[0008] A system self-learning module is used to periodically compare the predicted corrosion thickness reduction with the measured value of the ultrasonic thickness gauge, and automatically calibrate the corrosion rate model.

[0009] The technical effects and advantages of the present application are as follows: 1. The present application can quantify the corrosion impact of environmental emergencies, realize early warning, and avoid structural safety accidents caused by sudden corrosion events.

[0010] 2. The present application first incorporates real-time stress monitoring data into the corrosion life evaluation model, accurately depicts the coupling effect of stress corrosion, and greatly improves the accuracy of life prediction.

[0011] 3. The system of the present application models and predicts the unique environment and stress state of each specific structure, rather than using a general empirical model, and the results are more instructive.

[0012] 4. The present application realizes the transition from planned maintenance to condition-based maintenance, avoids waste caused by excessive maintenance, and prevents risks caused by insufficient maintenance, thereby significantly reducing the whole life cycle cost. BRIEF DESCRIPTION OF DRAWINGS

[0013] Figure 1 is a structural diagram of the present application.

[0014] Figure 2 is a flowchart of the present application. DETAILED DESCRIPTION

[0015] The technical solutions in the present application will be described clearly and completely below in combination with the drawings in the present application, and in addition, the forms of each structure described in the following embodiments are only examples, and all other embodiments obtained by those skilled in the art without making creative efforts belong to the scope of protection of the present application.

[0016] REFERENCE Figure 1 The present application provides a steel structure corrosion prevention evaluation system based on environmental data and AI, which comprises a data perception module, an environmental index generation module, an impact function generation module, a stress-corrosion coupling effect model construction module, a steel structure life prediction module, a grading early warning module and a system self-learning module.

[0017] REFERENCE Figure 2 The specific implementation steps of the present application include the following steps: S1, real-time collection of environmental data and steel structure state data of the steel structure.

[0018] It needs to be specifically pointed out that the collection of the environmental data of the steel structure specifically includes: Temperature and humidity sensor: monitor the atmospheric temperature and humidity to judge the condensation risk; Corrosive ion sensor: monitor the concentrations of chloride ions (Cl⁻) and sulfur dioxide (SO2); pH value sensor: monitor the acidity and alkalinity of the surface electrolyte; Weather station: monitor the wind speed, wind direction and rainfall for analyzing sudden environmental events.

[0019] The collection of the steel structure state data specifically includes: Strain gauge / fiber Bragg grating sensor (FBG): real-time monitoring of micro-strain and stress changes; Corrosion rate sensor: such as electrochemical noise (EN), linear polarization resistance (LPR), etc., directly measuring the instantaneous corrosion rate; Ultrasonic thickness gauge: periodically and automatically measuring the residual thickness of the base material under the coating.

[0020] S2, generating a real-time environmental corrosivity index through weighted calculation according to the current environment data of the steel structure.

[0021] It needs to be specifically pointed out that the real-time environmental corrosivity index is specifically: ; Where E is the environmental corrosivity index at the ith time point, which comprehensively calculates various complex environmental factors to quantify environmental corrosivity; the closer E is to 1, the stronger the corrosivity of the current environment to the steel structure; the closer to 0, the weaker the corrosivity.

[0022] j is an index variable, representing the jth environmental factor, m is the total number of environmental factors, and the summation calculation process traverses m environmental factors; For example: j=1 represents the chloride ion concentration (Cl⁻), j=2 represents the sulfur dioxide concentration (SO2), j=3 represents the relative humidity (RE), j=4 represents the temperature (T), and j=5 represents the pH value; then m=5.

[0023] W j is the weight of the jth environmental factor, between 0 and 1, and the sum of the weights of all factors is equal to 1; different environmental factors have different contributions to corrosion, and the weight W j is larger, indicating that the factor is more important in the comprehensive index, and its change has a greater impact on the final environmental corrosivity. Entropy weight method is used for calculation, which is determined by the dispersion degree of the data itself. The greater the data fluctuation of a factor, the more information it contains, and the system automatically gives it a higher weight.

[0024] X ij is the value of the jth environmental factor after standardization at the ith time point; standardization maps all factor values to 0 to 1, eliminating unit and order of magnitude differences, making them comparable; in corrosion analysis, the larger the value, the stronger the corrosion, called a positive indicator; the smaller the value, the stronger the corrosion, called a negative indicator. The standardization process includes conversion formulas to handle both cases, ensuring that the final X ij value is larger, indicating that the corrosion is stronger, and the logical direction is consistent.

[0025] S3, identify environmental mutations, combine the change intensity and duration of the environmental corrosivity index to build an environmental mutation impact function to calculate the additional corrosion amount.

[0026] It needs to be specifically pointed out that the system automatically identifies environmental mutation events (such as the start of acid rain, salt fog attack) through real-time monitoring of E value using the CUSUM change point detection method. Once a mutation is identified, the system will quantify the intensity ΔE and duration Δt of the event to build an environmental mutation impact function.

[0027] The environmental mutation impact function is specifically: ; Wherein C is the environmental mutation impact function, indicating the part of the corrosion amount caused by the environmental mutation event beyond the normal corrosion amount, for accumulation to the total corrosion amount, so as to make the life prediction more accurate; Alpha is an empirical constant, which is a calibration parameter of the model, determines the scale of the whole function, adjusts the numerical value output by the function to the range of actual observation value, needs to be determined by historical data fitting or laboratory experiment, the most suitable alpha value is deduced by analyzing the actual measured corrosion increment after multiple mutation events in the past; Delta E is the environmental corrosivity index change intensity, that is, the intensity of environmental mutation event, quantifies the intensity of the event, the larger the delta E, the more serious the environmental deterioration, the more intense the event, and the more additional corrosion caused, the calculation formula is: delta E = Es-Eb, Es is the peak value of E during mutation, Eb is the baseline value of the environment before mutation, taking the average value of the previous stage; Beta is a strength index, which is an empirical index, used to adjust the influence degree of intensity delta E on the final result, if beta > 1, it means that the influence of intensity is super-linear, that is, the additional corrosion will increase a lot when the intensity increases a little, which is determined by data fitting, reflecting the sensitivity of corrosion process to environmental intensity change; Delta t is the duration of environmental mutation event, that is, the time length experienced by E value from abnormal rise to fall to normal level, quantifies the persistence of the event, the longer the delta t, the longer the structure is exposed to harsh environment, and the more additional corrosion accumulated; Gamma is a duration index, which is an empirical index, used to adjust the influence degree of duration delta t on the final result, the relationship between corrosion amount and time is linear, that is, gamma = 1, which means that if the time is doubled, the corrosion amount will also be doubled, if gamma < 1, it means that there is a saturation effect, the influence of time is decreasing; If > 1, it means that the longer the time, the more serious the corrosion acceleration, because the fresh metal is continuously exposed after the corrosion products are washed away, which is determined by data fitting, reflecting the characteristics of corrosion process accumulation with time.

[0028] S4, according to the real-time state data of steel structure, real-time environmental corrosivity index and material properties, selecting algorithm and training to obtain stress-corrosion coupling effect model, combining with impact function to output actual corrosion rate.

[0029] It needs to be specifically pointed out that the construction steps of the stress-corrosion coupling effect model are specifically: A1, define input features, form the input vector X of the model through model influencing factors, specifically: X = [σ, f, E, M]; Where σ is the real-time stress value, static or quasi-static stress level is the main factor leading to stress corrosion cracking (SCC), the higher the stress, the stronger the atomic activity, the easier the protective film is broken, and the easier the corrosion is carried out; F is the stress change frequency, which is obtained by Fourier transform (FFT) time analysis of the stress signal, and alternating stress is the main reason leading to corrosion fatigue; E is the real-time environmental corrosion index, the same stress level produces different acceleration effects in different corrosive environments; M is the material attribute, including yield strength, chemical composition and heat treatment state, different materials have different resistance to stress corrosion, and these attributes are input to enable the model to distinguish different material structures and realize personalized prediction.

[0030] A2, select algorithm and train the algorithm; select gradient boosting tree GBRT, the relationship between stress, environment and corrosion rate is complex, GBRT can automatically capture these complex nonlinear interactions, after training, the model reflects which feature has greater influence, and has interpretability; The steps of the algorithm training are specifically: B1, collect a large amount of historical data, including input features X and corresponding stress corrosion factor S = R / Rb at each time point; Where S is the stress corrosion factor, indicating the amplification multiple of the current stress state on the basic corrosion rate; R is the actually measured corrosion rate, which is the real corrosion rate under the joint action of stress and environment; Rb is the corrosion rate caused only by the environment, which is a benchmark value, calculated by the current environmental corrosivity index; The stress corrosion factor quantifies the acceleration effect of stress, S = 1 indicates that stress does not accelerate corrosion, and corrosion is completely caused by the environment, S > 1 indicates that stress is accelerating corrosion, for example, S = 2.5 means that the current stress level makes the corrosion speed 2.5 times that of pure environmental action, the larger the S value, the more serious the stress corrosion coupling effect, and the more dangerous the structure; B2, input the data set (X, S) into the GBRT algorithm for training; B3, the algorithm minimizes the mean square error between the predicted value of the stress corrosion factor and the real value of the stress corrosion factor by continuously iterating, that is, adding a decision tree; B4, after training, a stress corrosion factor prediction model that can be put into use is obtained.

[0031] A3, model application, real-time prediction of actual corrosion rate Ra according to predicted stress corrosion factor, real-time environmental corrosion rate and impact function; The system packs the latest sensor data (σ, f, E, M) into a feature vector X every second / minute, inputs X into the trained GBRT model, and the model outputs the current Sa instantly. The system finally calculates the actual corrosion rate Ra = Re * Sa + C, where Re is the real-time environmental corrosion rate, Sa is the predicted stress corrosion factor, and C is the real-time impact function. If there is no environmental mutation at the current time, C = 0.

[0032] S5, calculate the current time steel structure corrosion cumulative damage according to the actual corrosion rate and predict the life.

[0033] It needs to be specifically pointed out that the cumulative damage is specifically: ; Where D(t) is the cumulative damage, the total damage amount accumulated by the steel structure due to corrosion from the time the steel structure is used, i.e. time 0 to the current time t, by monitoring the value of D(t) in real time to judge the health status of the structure; 0 is the starting point of integration, usually representing the time when the structure is put into use, or the initial time when the system starts monitoring, t is the end point of integration, representing the current time, defining the time range of cumulative damage calculation; by integrating the corrosion rate at every instant between time 0 and time t, the accurate total damage amount is obtained; Ra(τ) is the actual corrosion rate, i.e. the real-time corrosion rate at a certain specific instant τ, which is the integrand in the integral formula, considering the normal environment, stress acceleration effect and the impact of environmental mutation events; where τ is the integral variable, τ will take every time point between 0 and t, so that the corrosion rate at each point is included in the calculation, and finally after the integral is completed, τ itself will not appear in the result D(t).

[0034] The steps of the life prediction are specifically: C1, set the critical damage value Dc, i.e. the maximum corrosion damage that the structure can withstand before failure, which is determined by design specifications, laboratory experiments or historical data; C2, based on the current state D(t), input the future environmental prediction scenario and the future stress prediction scenario; C3, use the trained model to simulate and calculate the actual corrosion rate at each future time step by step, and calculate the cumulative future damage Df; C4, when Df ≥ Dc, the corresponding future time point is the predicted failure time T, and the remaining life = failure time - current time.

[0035] S6, combine the output results of each model and the sensor data to perform graded warning, and provide visual decision support.

[0036] Specifically, the hierarchical early warning is specifically: The first early warning is a long-term trend warning; the triggering condition specifically includes: Environmental trend: EC7>E30*1.3, that is, the average environmental corrosion index in the past 7 days is more than 1.3 times the average index in the past 30 days, indicating that the environmental corrosion has a clear deterioration trend, not just a single day fluctuation; Damage progress: D(t) / Dc>0.4, that is, the cumulative damage has reached 40% of the critical damage value, which is a macroscopic early warning based on the design life progress, indicating that the structure has passed the initial stage; System response: log recording, the structure status indicator light in the visual board changes to yellow. The system automatically generates a weekly report and lists the structure in the focus list.

[0037] The second early warning is a short-term time warning; the triggering condition specifically includes: Environmental mutation: ΔE>E30*0.8 and Δt>2 hours, the intensity of a single environmental mutation event is more than 0.8 times the average ECI in the past 30 days, and the duration of the event is more than 2 hours; Stress surge: S(t) / SCF24h>1.8 and S(t)>2.0, that is, the current stress corrosion factor has soared by 80% compared to the average value in the past 24 hours, and its absolute value is greater than 2.0, meaning that the stress level is not only high, but also has changed abnormally in the short term, and the corrosion is being rapidly accelerated; Lifetime warning: T<Ts, that is, the remaining life predicted by the system is shorter than the safety margin required by the design specification Ts; System response: automatically send warning emails / text messages to the operation team leader, the structure icon in the visual board interface changes to orange and slowly flashes, the system automatically generates a diagnosis report, indicating that the main cause is an environmental event or a stress anomaly.

[0038] The third alarm is an immediate emergency warning, and the triggering condition specifically includes: Damage threshold D(t) / Dc>0.9, that is, the cumulative damage has exceeded 90% of the critical value, and the structure is at the edge of failure, with high risk; Crack initiation: acoustic emission sensors monitor high-frequency acoustic emission signals (>100 times / second) and high signal strength (>80 decibels), which are typical characteristics of micro-cracks rapidly expanding and the most direct precursor of macroscopic fracture of the structure; Structural response: the real-time monitored strain value exceeds 80% of the material yield strain, indicating that the structure has entered the nonlinear stage and local plastic deformation may be occurring, with a significant decrease in carrying capacity.

[0039] System response: Trigger audible and visual alarms, automatically call the on-duty room, and pop up the highest priority alert on the visual board interface. The system immediately locks the data and generates an emergency report, strongly suggesting immediate on-site disposal and safety assessment.

[0040] The visual decision support specifically includes: Global overview map: A geographic information map with all monitored steel structures marked. The real-time health status of each structure is intuitively displayed through yellow / orange / red colors. Clicking on any structure allows you to drill down to view detailed data.

[0041] Real-time data panel: Digital and instrument panel displays current key indicators such as E, S, Ra, and maximum stress values in real time; trend curve displays the historical curve of key indicators over time.

[0042] Remaining life prediction curve: A chart predicting future time, with X-axis as time and Y-axis as cumulative damage D(t). The chart includes historical damage curve, predicted damage curve, critical line, and predicted failure point. The historical damage curve is the actual D(t) trajectory from the past to the present, the predicted damage curve is the growth path of future D(t) based on current environmental trends, the critical line is a horizontal line marked with Dc, and the predicted failure point is the intersection of the predicted curve and the critical line, marked with the predicted date.

[0043] Corrosion risk heat map: Based on the corrosion rate or remaining life data of each sensor location, a color gradient heat map is generated at the corresponding location of the structure model, with blue representing low risk and red representing high risk.

[0044] Maintenance action recommendation list: An automatically generated to-do list, with specific and executable recommendations pushed by the system based on risk positioning and diagnosis results.

[0045] S7. Regularly calculate the predicted corrosion thickness reduction and compare it with the measured value of the ultrasonic thickness gauge to automatically calibrate the corrosion rate model.

[0046] Specifically, the step of automatically calibrating the corrosion rate model is: D1. Select the calibration time period, select the last ultrasonic measurement time t1 to the current ultrasonic measurement time t2; obtain the measured data, the thickness value Tt1 measured by ultrasonic at time t1, and the latest thickness value Tt2 measured by ultrasonic at time t2; calculate the actual thickness reduction: ΔD = Tt1 - Tt2; D2. Obtain the predicted data, extract the time series data of the instantaneous corrosion rate Ra(τ) predicted by the model from t1 to t2 from the database, and calculate the predicted thickness reduction: ; accomplished by summation of sampling intervals (e.g. one data per minute), ADp ~∑[Ra(ti)*At], where At is the sampling time interval; D3, calculate absolute error and relative error and analyze; absolute error A = ADp- AD, relative error B = A / AD, where ADp is the predicted corrosion thickness reduction, AD is the actual corrosion thickness reduction; if B > 0, the model prediction is too conservative, the predicted corrosion amount is more than the actual measured; if B < 0, the model prediction is too aggressive, underestimates the actual corrosion amount; if |B| < 0.1, the model accuracy is higher, the error is within 10%, no adjustment or fine-tuning is needed.

[0047] D4, model calibration, the system mainly calibrates the output scale of the environmental benchmark corrosion rate model Re and the stress corrosion factor S; A calibration factor K is introduced to modify the entire corrosion rate model to calculate a new calibration factor: Knew = AD / ADp, where ADp is the predicted corrosion thickness reduction, AD is the actual corrosion thickness reduction; the new calibration factor is applied to future predictions, Ra'(τ) = K*Ra, the initial K = 1, after this calibration, update K = 0.7*Kold + 0.3*Knew, Kold is the previous K value, Knew is the new K value, smooth update is adopted to avoid sharp fluctuations caused by single error.

[0048] D5, write the new calibration factor K into the configuration database of the system, for all subsequent prediction calculations, record all information of this calibration: calibration time t2, time period At = t2-t1, AD, ADp, A, B, old Kold value, new Knew value and updated K value.

[0049] It should be noted that if |B| exceeds 25% for two consecutive times, indicating that a sensor fails or the model deviates completely, the system should trigger a warning for the health of the system itself, and notify the engineer to perform manual inspection.

[0050] From the above description of the embodiments, those skilled in the art can clearly understand that the embodiments of the present application can be realized by means of software or software combined with necessary general hardware platforms, and of course can also be realized by hardware functions. Based on such understanding, the technical solutions of the present application can be embodied in the form of a software product, which is stored in a storage medium, includes a plurality of instructions for causing a computer device, such as but not limited to a personal computer, a server, or a network device, to execute all or part of the steps of the method described in any embodiment of the present application.

[0051] The above describes the exemplary embodiments of the present application, it should be understood that the above exemplary embodiments are not restrictive, but illustrative, the protection scope of the present application is not limited thereto, it should be understood that the person skilled in the art can modify and change the embodiments of the present application without departing from the spirit and scope of the present application, and these modifications and changes should be within the protection scope of the present application.

Claims

1. A steel structure corrosion assessment system based on environmental data and AI, characterized in that, Specifically, it includes: Data sensing module: used to collect real-time environmental data and steel structure status data of the steel structure; Environmental index generation module: used to generate a real-time environmental corrosion index based on the current environmental data of the steel structure through weighted calculation; Impact function generation module: used to identify environmental abrupt changes, and combine the intensity and duration of changes in the environmental corrosivity index to construct an environmental abrupt change impact function to calculate the additional corrosion amount; Stress-corrosion coupling effect model construction module: It is used to select an algorithm and train it to obtain a stress-corrosion coupling effect model based on the real-time status data of steel structure, real-time environmental corrosion index and material properties, and output the actual corrosion rate by combining the impact function. Steel structure life prediction module: used to calculate the cumulative corrosion damage of steel structure at the current moment based on the actual corrosion rate and to predict its life. The tiered early warning module combines the output results of various models with sensor data to provide tiered early warnings and offers visual decision support. System self-learning module: used to periodically calculate and compare the predicted corrosion thickness reduction with the measured value of the ultrasonic thickness gauge, and automatically calibrate the corrosion rate model.

2. The steel structure corrosion assessment system based on environmental data and AI according to claim 1, characterized in that: The environmental data of the steel structure specifically includes: Atmospheric temperature and humidity, chloride ion and sulfur dioxide concentrations, acidity and alkalinity of surface electrolyte, wind speed, wind direction and rainfall; specific steel structure condition data include: strain and stress changes, instantaneous corrosion rate and remaining thickness of the base material under the coating.

3. The steel structure corrosion assessment system based on environmental data and AI according to claim 1, characterized in that: The real-time environmental corrosivity index is specifically: ； Where E is the environmental corrosivity index at the i-th time point; j is the index variable, representing the j-th environmental factor; m is the total number of environmental factors, indicating that the summation calculation process will traverse m environmental factors; W j The weight of the j-th environmental factor is calculated using the entropy weight method. The weights of each environmental factor are between 0 and 1, and the sum of the weights of all factors equals 1. ij Let be the standardized value of the j-th environmental factor at time point i.

4. The steel structure corrosion assessment system based on environmental data and AI according to claim 1, characterized in that: The specific environmental abrupt change shock function is as follows: ； Where C is the environmental abrupt change impact function, representing the portion of corrosion exceeding the normal corrosion level caused by the environmental abrupt change event; α is the model calibration parameter, derived by back-calculating the corrosion increment measured after multiple past abrupt change events; ΔE is the intensity of the change in the environmental corrosivity index, calculated as: ΔE=Es-Eb, where Es is the peak value of E during the abrupt change, and Eb is the baseline value of the environment before the abrupt change, taking the average value of the previous stage; β is the intensity index, determined through data fitting; Δt is the duration of the environmental abrupt change event, i.e., the time it takes for the E value to rise abnormally and fall back to the normal level; γ is the duration index, determined through data fitting.

5. The steel structure corrosion assessment system based on environmental data and AI according to claim 1, characterized in that: The specific steps for constructing the stress-corrosion coupling effect model are as follows: A1. Define the input features. The input vector X of the model is constructed by the influencing factors of the model. Specifically: X=[σ, f, E, M]; where σ is the real-time stress value, F is the stress change frequency, E is the real-time environmental corrosion index, and M is the material property. A2. Select an algorithm and train it. Choose Gradient Boosting Tree (GBRT) to capture complex nonlinear interactions. The trained model can be used to predict stress corrosion factor. A3. Model Application: Based on the predicted stress corrosion factor, real-time environmental corrosion rate, and impact function, the actual corrosion rate Ra is predicted in real time. Ra = Re * Sa + C, where Re is the real-time environmental corrosion rate, calculated from the real-time environmental corrosivity index, Sa is the predicted stress corrosion factor, and C is the real-time impact function. If no sudden environmental change occurs at the current moment, then C = 0.

6. The steel structure corrosion assessment system based on environmental data and AI according to claim 5, characterized in that: The specific steps for training the algorithm are as follows: B1. Collect a large amount of historical data, including the input feature X and the corresponding stress corrosion factor S=R / Rb at each time point; where S is the stress corrosion factor, R is the actual measured corrosion rate, and Rb is the corrosion rate caused only by the environment, which is calculated by the current real-time environmental corrosion index. B2. Input the dataset (X, S) into the GBRT algorithm for training; B3. The algorithm minimizes the mean square error between the predicted and actual values ​​of stress corrosion factor through continuous iteration. B4. After training is complete, the model can predict stress corrosion factor.

7. The steel structure corrosion assessment system based on environmental data and AI according to claim 1, characterized in that: The cumulative damage specifically refers to: ； Where D(t) is the cumulative damage, which is the total amount of damage accumulated by the steel structure due to corrosion from the time when the steel structure was put into use, i.e., time 0, to the current time t; 0 represents the time when the structure was put into use, and t represents the current time; Ra(τ) is the actual corrosion rate, and τ is the integration variable, which takes over every time point between 0 and t, and the corrosion rate at each point is included in the calculation.

8. The steel structure corrosion assessment system based on environmental data and AI according to claim 1, characterized in that: The specific steps for lifetime prediction are as follows: C1. Set the critical damage value Dc, which is the maximum corrosion damage that the structure can withstand before failure. It is determined by design specifications, laboratory experiments, or historical data. C2. Based on the current state D(t), input the future environmental prediction scenario and the future stress prediction scenario. C3. Using the trained model, simulate and calculate the actual corrosion rate at each future time step by step, and calculate the cumulative future damage Df; C4. When Df≥Dc, the corresponding future time point is the predicted failure time T, and the remaining lifetime = failure time - current time.

9. The steel structure corrosion assessment system based on environmental data and AI according to claim 1, characterized in that: The tiered early warning system specifically refers to: Level 1 warning is a long-term trend warning; the specific triggering conditions include: environmental trend: EC7 > E30*1.3, that is, the average environmental corrosion index in the past 7 days exceeds 1.3 times the average index in the past 30 days; damage progress: D(t) / Dc > 0.4, that is, the cumulative damage has reached 40% of the critical damage value; Level 2 warning is a short-term, time-sensitive warning. The specific triggering conditions include: Environmental abrupt change: ΔE > E30*0.8 and Δt > 2 hours, meaning the intensity of a single environmental abrupt change event exceeds 0.8 times the 30-day average ECI, and the event lasts for more than 2 hours; Stress surge: S(t) / SCF24h > 1.8 and S(t) > 2.0, meaning the current stress corrosion factor has surged by 80% compared to the average of the past 24 hours, and its absolute value is greater than 2.0; Life warning: T < Ts, meaning the system's predicted remaining life is shorter than the safety margin Ts required by the design specifications. Level 3 early warning is an immediate and urgent warning. The specific triggering conditions include: damage threshold D(t) / Dc > 0.9, meaning that the cumulative damage has exceeded 90% of the threshold value; crack initiation: the acoustic emission sensor detects high-frequency acoustic emission signals (>100 times / second) with high signal intensity (>80 dB); structural response: the strain value monitored in real time exceeds 80% of the material's yield strain. This indicates that the structure has entered the nonlinear stage.

Images