Visual drive based organic coating corrosion state assessment and maintenance decision method

Abstract

The present application belongs to the field of material corrosion protection and intelligent operation and maintenance, and relates to a kind of organic coating corrosion state evaluation and maintenance decision method based on visual driving.The method obtains organic coating surface image data, combines the corrosion grade information output by the visual recognition model, and introduces the electrochemical calibration results and environmental factors, and quantitatively evaluates the corrosion state through a comprehensive corrosion index calculation model, establishes the mapping relationship between the corrosion grade and the maintenance strategy, and dynamically adjusts the corrosion development trend to realize the quantitative evaluation of the coating state and the output of the maintenance decision.The method can convert the corrosion recognition results into specific engineering measures, improve the scientificity and consistency of the maintenance decision, and is suitable for intelligent operation and maintenance management of various metal structure organic coatings.

Description

Technical Field

[0001] This invention relates to the field of intelligent assessment and maintenance decision-making technology for material corrosion, and particularly to a vision-driven method for assessing the corrosion status and making maintenance decisions for organic coatings. Background Technology

[0002] Organic coatings are the most important and economical means of corrosion protection for metal structures. However, with increasing service life, coatings inevitably age, blister, crack, and even peel off, ultimately leading to corrosion of the base metal and threatening structural safety. Therefore, accurate assessment of coating corrosion status and timely maintenance measures are crucial for ensuring equipment availability and extending service life.

[0003] In actual engineering operation and maintenance, corrosion assessment and maintenance decisions typically rely on human experience. Maintenance personnel make qualitative judgments about the degree of corrosion through visual inspection, and then decide whether treatment is necessary and what treatment measures to take based on their personal experience. This model has significant drawbacks: First, assessment standards are not uniform, and different personnel may have significantly different judgments on the same corrosion state, leading to poor decision-making consistency. Second, qualitative judgments cannot quantify the development trend of corrosion, easily resulting in over-maintenance or under-maintenance. Over-maintenance not only wastes human and material resources but may also cause unnecessary damage to intact coatings; under-maintenance may lead to further corrosion and create greater safety hazards. In addition, although some existing intelligent identification technologies can identify corrosion levels, the output is only a level label, lacking connection with subsequent maintenance measures. There is a significant disconnect between detection and operation and maintenance, and it cannot directly form a closed-loop decision support.

[0004] Therefore, there is an urgent need to establish a complete methodology from corrosion identification to maintenance decision-making, which can organically integrate visual identification results, electrochemical calibration data and environmental influencing factors to generate quantitative corrosion status assessment values, and output clear and executable maintenance strategies based on scientific decision-making rules, thereby achieving intelligent and refined management of corrosion protection. Summary of the Invention

[0005] The purpose of this invention is to provide a vision-driven method for assessing the corrosion status of organic coatings and making maintenance decisions. This method aims to overcome the limitations of existing technologies, such as the disconnect between identification and decision-making and reliance on human experience. By constructing a closed-loop method of "identification result → status assessment (including comprehensive corrosion index calculation) → dynamic decision-making", the method achieves automation of corrosion assessment and scientification of maintenance decisions.

[0006] To achieve the above objectives, the core of the technical solution proposed in this invention lies in: using the corrosion level output by visual recognition as basic information, physically calibrating the visual level using electrochemical calibration results, then dynamically correcting it by introducing environmental factors, obtaining a quantified corrosion state value through a comprehensive corrosion index calculation model, then dynamically evaluating it in conjunction with the corrosion development trend, and finally automatically outputting targeted maintenance suggestions based on a preset corrosion level-maintenance strategy mapping relationship. Specifically, this invention is achieved through the following technical solution: The first step is to acquire image data of the organic coating surface. Using a portable image acquisition terminal or a fixed monitoring device, photograph the coating area to be evaluated to obtain high-resolution digital images. Image acquisition should ensure basic illumination uniformity; a standard light source may be used if necessary.

[0007] The second step involves analyzing the image based on a visual recognition model to obtain a preliminary corrosion level. The acquired image is input into a pre-trained high-precision corrosion visual recognition model (such as the electrochemically constrained corrosion visual recognition model disclosed in the inventor's prior patent). The model calculates and outputs the corrosion level corresponding to the image. This level is typically divided into at least four levels: Level 0 (intact), Level 1 (slight aging / early stage of water seepage), Level 2 (intermediate failure / blistering), Level 3 (severe corrosion / localized peeling), and Level 4 (corrosion of the base metal).

[0008] The third step involves a quantitative assessment of the corrosion state based on the electrochemical calibration results. While the corrosion level output by the visual recognition model has high accuracy, misjudgments may still occur due to factors such as surface contamination and abnormal lighting. Therefore, this invention introduces electrochemical calibration results as the calibration basis. The electrochemical calibration results can be the low-frequency impedance modulus |Z|0.01Hz measured at the same measurement point using a portable electrochemical workstation, or a correlation between the electrochemical parameters of the coating system and the corrosion level established using historical data. By comparing the visual level with the electrochemical parameters, if there is a significant difference in the degree of corrosion indicated by the two, the visual level is appropriately adjusted to obtain a calibrated corrosion level Lc.

[0009] The fourth step involves incorporating environmental factors to correct the assessment results and calculating the comprehensive corrosion index (CCI). In actual service environments, environmental factors such as temperature, humidity, and salt spray deposition directly affect the coating's degradation rate and corrosion development trend. Therefore, after obtaining the calibrated corrosion level, it is necessary to make corrections based on the environmental conditions of the measurement points. This invention establishes an environmental factor correction function f(E), incorporating environmental parameters (such as annual average humidity and salt spray deposition) into the comprehensive assessment model to weight and adjust the calibrated level, ultimately obtaining a quantified comprehensive corrosion index (CCI). The CCI calculation formula is: CCI = Lc × f(E), where f(E) is the environmental correction function, typically in the form f(E) = 1 + α•(RH - RH0) + β•S, where RH is the relative humidity, S is the salt spray deposition rate, RH0 is the baseline humidity, and α and β are coefficients obtained by fitting historical corrosion rate data. CCI is a continuous value, which can more precisely reflect the severity and development trend of the corrosion state.

[0010] The fifth step is to conduct a dynamic assessment based on the corrosion development trend. To guide maintenance decisions more scientifically, this invention introduces a dynamic assessment mechanism. If historical detection data exists for the same measuring point, the corrosion rate v = (CCI2 - CCI1) / Δt can be calculated, characterizing the speed of corrosion development. When the corrosion rate v exceeds a preset threshold vth, it indicates that the coating is in an accelerated deterioration stage, and even if the current CCI value is not high, the urgency of maintenance should be increased. The results of the dynamic assessment will be used to adjust the subsequent decision level.

[0011] Step 6: Based on the mapping relationship between corrosion state and preset maintenance strategy, output the corresponding maintenance decision. A maintenance strategy mapping rule base based on CCI values ​​is pre-established based on extensive engineering experience and standard specifications. The rule base contains the following correspondences: When CCI ≤ T1, output "No processing required, continue monitoring"; When T1 < CCI ≤ T2, output "Regular monitoring, it is recommended to re-examine every six months"; When T2 < CCI ≤ T3, output "Local repair, polishing and repainting recommended"; When T3 < CCI ≤ T4, output "Partial recoating, it is recommended to remove the failed coating and recoat"; when CCI > T4, output "Full recoating, it is recommended to thoroughly remove rust and prevent corrosion of the entire structure".

[0012] If the dynamic assessment step identifies a corrosion rate v > vth, then based on the decision output by the above rules, the decision level is raised by one level (e.g., from "periodic monitoring" to "local repair") to reflect the urgency of accelerated corrosion. The output decision information may include written suggestions, process guidance, consumables lists, etc., which are directly used to guide on-site maintenance operations.

[0013] The beneficial effects of this invention are as follows: First, by introducing electrochemical calibration and environmental factor correction, the physical calibration and dynamic adjustment of visual recognition results are achieved, significantly improving the accuracy and reliability of corrosion status assessment. Second, through the quantitative calculation of the Comprehensive Corrosion Index (CCI), multi-source information is integrated into a comparable numerical indicator, providing a quantitative basis for scientific decision-making. Third, the introduction of a dynamic evaluation mechanism ensures that decisions not only rely on the current state but also consider corrosion development trends, avoiding static decision-making based on outdated methods. Fourth, a quantitative mapping relationship between corrosion status and maintenance strategies is established, automatically transforming evaluation results into specific and executable engineering measures, completely solving the problem of disconnect between detection and operation and maintenance. Fifth, this method achieves standardization and automation of maintenance decisions, avoiding the subjectivity and inconsistency brought about by human experience, which helps optimize the allocation of maintenance resources and reduce the total life-cycle maintenance cost. Finally, this method has a complete system, strong operability, and is easy to integrate into existing equipment management information systems or intelligent inspection terminals, providing key technical support for building an intelligent and digital corrosion protection system. Attached Figure Description

[0014] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments: Figure 1 This is a flowchart of the overall process of the vision-driven organic coating corrosion status assessment and maintenance decision-making method of the present invention, which shows the complete technical route from image acquisition, visual recognition, electrochemical calibration, environmental correction to decision output.

[0015] Figure 2 This is a schematic diagram illustrating the mapping relationship between corrosion level and maintenance strategy in this invention, exemplarily showing typical maintenance recommendations corresponding to different corrosion levels.

[0016] Figure 3 This is a system schematic diagram of the present invention in the application of steel structure tower engineering, showing the process of on-site data acquisition, background processing and decision output. Detailed Implementation

[0017] The present invention will be further explained below with reference to specific implementation schemes, but it is not limited to the present invention. The structures, proportions, sizes, etc. shown in the accompanying drawings are only used to complement the content disclosed in the specification, so as to enable those skilled in the art to understand and read, and are not intended to limit the conditions under which the present invention can be implemented. Therefore, they have no substantial technical significance. Any modification of the structure, change of the proportion relationship or adjustment of the size, without affecting the effect and purpose that the present invention can produce, should still fall within the scope of the technical content disclosed in the present invention.

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to two specific embodiments. Embodiment 1 will elaborate on how, in a laboratory environment, known corrosion samples and simulation data are used to verify the logical correctness of the state assessment and decision mapping in the method of this invention. Embodiment 2 will demonstrate the actual engineering application of this method on a large steel structure bridge in a coastal industrial area in northern my country, comprehensively evaluating its on-site effects.

[0019] Example 1: Method Validation in a Laboratory Environment Preparation of experimental materials and data.

[0020] In this embodiment, Q235 carbon steel was used as the substrate, and 30 coating samples of 150mm×75mm×3mm were prepared. The coating system consisted of epoxy zinc-rich primer (60μm) + epoxy micaceous iron oxide intermediate coat (80μm) + aliphatic polyurethane topcoat (60μm). By controlling the duration of the accelerated corrosion test, a batch of samples covering different corrosion states from level 0 to level 4 were obtained. For each sample, surface images were first taken using an industrial camera under standard light source, and then the 0.01Hz low-frequency impedance modulus |Z|0.01Hz was measured using an electrochemical workstation. The corrosion level was independently assessed by three senior corrosion engineers according to ISO 4628 standard, and the majority opinion was taken as the true level. Finally, a validation set containing 30 sets of data (images, impedance values, and true levels) was constructed.

[0021] Visual identification and preliminary corrosion level.

[0022] The collected images were input into the electrochemically constrained visual recognition model trained by the inventors in their prior patent (Patent 3). The model outputs a preliminary corrosion level for each sample. The results show that the level output by the model is basically consistent with the actual level assessed by the engineer, but there are deviations in the output for 3 samples.

[0023] Electrochemical calibration.

[0024] Electrochemical impedance spectroscopy data was used to calibrate the biased samples. The known electrochemical calibration rule for this coating system is: |Z| 0.01 Hz > 5 × 10⁻⁶.8 Ω•cm 2 Corresponding to level 0, 10 7 ~5×10 8 Ω•cm 2 Corresponding to level 1, 10 5 ~10 7 Ω•cm 2 Corresponding to level 2, 10 3 ~10 5 Ω•cm 2 Corresponding to level 3, <10 3 Ω•cm 2 This corresponds to level 4. After calibration, the level of all samples is consistent with the level indicated by the electrochemical parameters, resulting in the calibration level Lc.

[0025] Environmental factor correction and comprehensive corrosion index calculation.

[0026] This embodiment was conducted under constant temperature and humidity conditions in a laboratory, where environmental factors remained essentially constant. To demonstrate the method, we simulated two environmental conditions: for samples #01 to #10, it was assumed that they were used in a mild indoor environment, and the environmental correction function f(E) was set to 0.9; for samples #11 to #30, it was assumed that they were used in an outdoor industrial atmospheric environment, and f(E) was set to 1.2. The comprehensive corrosion index CCI was calculated as: CCI = Lc × f(E). For example, the CCI of sample #07 (Level 2, mild indoor environment) was 1.8; the CCI of sample #15 (Level 3, industrial environment) was 3.6.

[0027] Dynamic evaluation (based on historical data).

[0028] In this embodiment, some samples have two sets of test data, which can be used to calculate the corrosion rate. For example, sample #21 had a CCI of 2.5 three months ago and a current CCI of 3.6. Therefore, the corrosion rate v = (3.6 - 2.5) / 3 ≈ 0.37 / month, which is higher than the preset threshold of 0.2 / month. Thus, the urgency needs to be increased when making decisions.

[0029] Maintain decision output.

[0030] The preset maintenance strategy mapping rules are based on CCI thresholds: T1=1.0, T2=2.0, T3=3.0, T4=4.0. For sample #07 (CCI=1.8), there is no dynamic improvement, and the output is "Regular monitoring, re-inspect every six months". For sample #15 (CCI=3.6), there is no dynamic improvement, and the output is "Local recoating, remove the failed coating and recoat". For sample #21 (current CCI=3.6, but v>vth), the output is increased by one level from "local recoating", and "Full recoating, recommended to be done during the next major anti-corrosion overhaul of the bridge". This result is completely consistent with the engineer's experience-based recommendations. This embodiment verifies the logical correctness and effectiveness of the method of the present invention.

[0031] Example 2: Application of large steel structure bridge in a coastal industrial zone in northern China Engineering background and challenges.

[0032] This embodiment selects a 580-meter main span cable-stayed bridge in a coastal industrial zone in northern my country as the engineering application object. Located in the Bohai Bay, the bridge experiences a typical warm temperate semi-humid continental monsoon climate, characterized by cold winters and hot, humid summers, and is constantly affected by industrial air pollution and marine salt spray. The bridge's steel structure uses an epoxy zinc-rich / epoxy micaceous iron oxide / polyurethane coating system and has been in service for eight years. In recent years, maintenance personnel have discovered varying degrees of aging in some areas, necessitating a comprehensive corrosion assessment and the development of a scientific and economical maintenance plan.

[0033] Data acquisition and processing.

[0034] The project team employed a combination of drones equipped with multiple sensors and aerial work platforms to select 50 representative monitoring points on key components of the bridge, including the towers, main beams, and crossbeams. At each monitoring point, the following data was collected: ① Five surface images were taken using a high-definition camera (equipped with a ring LED supplemental light) mounted on the drone; ② Environmental parameters (temperature, relative humidity, and salt spray deposition rate) were simultaneously recorded at the monitoring point using miniature temperature and humidity sensors and salt spray sensors attached to the drone; ③ Inspectors arrived at the monitoring point via an aerial work platform and used a portable solid gel electrolyte probe to measure the electrochemical impedance spectroscopy (EIS) at that point, obtaining the low-frequency impedance modulus. The entire data collection process lasted two weeks, covering typical weather conditions in spring and summer. Simultaneously, inspection data from some monitoring points on the bridge from one year prior was retrieved for corrosion rate calculations.

[0035] Status assessment and decision generation.

[0036] The collected image data is input into a visual recognition model, which outputs the preliminary corrosion level for each measurement point. Then, the calibration rules are calibrated using measured electrochemical impedance spectroscopy values. The electrochemical calibration threshold for the industrial area where the bridge is located, corrected from previous laboratory tests, is: |Z| 0.01Hz > 1×10⁻⁶.9 Ω•cm 2 Corresponding to level 0, 10 8 ~10 9 Corresponding to level 1, 10 6 ~10 8 Corresponding to level 2, 10 4 ~10 6 Corresponding to level 3, <10 4 Corresponding to level 4. After calibration, Lc is obtained.

[0037] Comprehensive corrosion index calculation. Based on measured environmental parameters, the environmental correction function is defined as f(E) = 1 + 0.01×(RH-60) + 0.05×S, where RH is the relative humidity (%) and S is the salt spray deposition rate (mg / 100cm³). 2 •h). This formula is derived based on historical corrosion rate data. After calculating the environmental factors for each measuring point, the comprehensive corrosion index CCI = Lc × f(E) is obtained. For example, for a measuring point located at the bottom of the main beam (Lc=2, RH=80%, S=1.2), f(E)=1+0.01×20+0.05×1.2=1.26, CCI=2.52; for a measuring point located at the top of the tower (Lc=3, RH=65%, S=0.8), f(E)=1+0.01×5+0.05×0.8=1.09, CCI=3.27.

[0038] Dynamic evaluation. For measuring points with historical data, calculate the corrosion rate v = (CCInow - CCIlast) / Δt (month). Set a threshold vth = 0.2 / month. For example, a measuring point on the lower flange of the main beam had CCIlast = 1.8 a year ago, and currently has CCInow = 2.52, v = 0.06 / month, which is below the threshold, so the decision level is not increased. Similarly, a measuring point at the crossbeam connection had CCIlast = 2.8 a year ago, and currently has CCInow = 4.72, v = 0.16 / month, which is also below the threshold, so the decision level is not increased either. However, if any measuring point's v exceeds the threshold, the decision level needs to be increased.

[0039] Maintenance decision output. Based on the CCI thresholds (T1=1.0, T2=2.0, T3=3.0, T4=4.0) and dynamic evaluation results, maintenance recommendations are generated.

[0040] Some of the results are shown in the table below: Engineering Effectiveness Evaluation. The maintenance recommendations output by this invention were compared with the independent recommendations given by three senior corrosion prevention engineers originally hired by the bridge management unit. There were inconsistencies among the three engineers regarding the recommendations for 50 monitoring points at 8 points, mainly concentrated in the boundary area between Level 2 and Level 3. The recommendations given by the method of this invention showed a high degree of consistency, and the agreement rate with the final consensus recommendations of most experts reached 94%. For the dynamically improving Zone D, the engineers originally recommended "regular monitoring," but this invention recommended "local repair" based on the high corrosion rate. Subsequent detailed testing of this area revealed signs of accelerated deterioration, verifying the effectiveness of the dynamic assessment. Referring to the evaluation report of this invention, the bridge management unit formulated a graded maintenance plan: lightly corroded points such as Zone A were included in routine inspections; moderately corroded points such as Zone B were scheduled for local repair by a professional team within two weeks; severely corroded points in Zone C were included as key areas for the bridge's major corrosion prevention overhaul next year; and accelerated deterioration points in Zone D were immediately scheduled for detailed testing and repair preparation. It is estimated that compared with the full-coverage, large-area recoating strategy used in previous years, the precise decision-making method of this invention directly saves approximately 35% of maintenance costs and avoids unnecessary sanding that damages the intact coating. Furthermore, by tracking the same batch of test points for two consecutive years, it was found that the corrosion recurrence rate in areas maintained according to the recommendations of this invention was significantly lower than in areas previously maintained based on experience.

[0041] Summary of project results.

[0042] The successful application of this embodiment on a large steel structure bridge fully demonstrates that the method of this invention can transform corrosion identification results into scientific, specific, and quantifiable maintenance decisions, effectively solving the engineering problems of inconsistent human experience judgments and the disconnect between detection and operation and maintenance. By introducing a comprehensive corrosion index calculation and dynamic evaluation mechanism, the evaluation results are closer to the actual state of the coating, and the decision recommendations are more targeted and economical. This method provides a mature technical solution for the intelligent operation and maintenance of large infrastructure such as bridges, port machinery, and offshore platforms, and has broad prospects for promotion and application.

[0043] In summary, this invention provides a vision-driven method for assessing the corrosion status of organic coatings and making maintenance decisions. By organically combining visual recognition, electrochemical calibration, environmental correction, comprehensive corrosion index calculation, and dynamic evaluation, it achieves closed-loop intelligent management from corrosion detection to maintenance decision-making, significantly improving the scientific level of corrosion protection. Matters not covered in this invention are prior art.

[0044] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A vision-driven method for assessing the corrosion status of organic coatings and making maintenance decisions, characterized in that, Includes the following steps: (1) Obtain image data of the organic coating surface; (2) The image is analyzed based on a visual recognition model to obtain the corrosion level; (3) Based on the electrochemical calibration results, a quantitative assessment of the corrosion state is performed; (4) Introduce environmental factors to revise the assessment results; (5) Output the corresponding maintenance decision based on the mapping relationship between corrosion state and preset maintenance strategy; The corrosion level is output by a visual recognition model.

2. The vision-driven organic coating corrosion status assessment and maintenance decision-making method according to claim 1, characterized in that, The electrochemical calibration results are used to calibrate the corrosion level.

3. The vision-driven organic coating corrosion status assessment and maintenance decision-making method according to claim 1, characterized in that, The environmental factors include humidity, salt spray, and temperature.

4. The vision-driven organic coating corrosion status assessment and maintenance decision-making method according to claim 1, characterized in that, The maintenance strategy includes regular monitoring, local repairs, and full recoating.

5. The vision-driven organic coating corrosion status assessment and maintenance decision-making method according to claim 1, characterized in that, The mapping relationship is established based on the corrosion level and development trend.

6. The vision-driven organic coating corrosion status assessment and maintenance decision-making method according to claim 1, characterized in that, The output includes corrosion status assessment values ​​and maintenance recommendations.

7. The vision-driven organic coating corrosion status assessment and maintenance decision-making method according to claim 1, characterized in that, Suitable for different types of organic coating systems.

8. The vision-driven organic coating corrosion status assessment and maintenance decision-making method according to claim 1, characterized in that, The quantitative assessment and correction described in steps (3) and (4) are achieved by calculating the comprehensive corrosion index. The specific calculation formula is: CCI = Lc × f(E), where Lc is the corrosion level after electrochemical calibration, f(E) is the environmental correction function constructed based on environmental parameters, and the output CCI value is used to characterize the severity of corrosion.

9. The vision-driven organic coating corrosion status assessment and maintenance decision-making method according to claim 8, characterized in that, The environmental correction function f(E) includes at least one environmental parameter, and its expression is f(E) = 1 + α•(RH-RH0) + β•S, where RH is the relative humidity, S is the salt spray deposition rate, and α and β are coefficients obtained by fitting historical data.

10. The vision-driven organic coating corrosion status assessment and maintenance decision-making method according to claim 1, characterized in that, The maintenance decision generation in step (5) includes a dynamic evaluation mechanism: the corrosion rate v is calculated based on historical detection data. When v exceeds a preset threshold, the decision level is raised by one level based on the maintenance strategy determined according to the current CCI value, so as to reflect the impact of corrosion development trend on maintenance urgency.

Images