Method and system for detecting performance of coastal steel structure

Abstract

The invention relates to the technical field of steel structure detection, and discloses a coastal steel structure performance detection method and system, and the method comprises the steps: carrying out the clustering sequence division of all pixel points based on each pixel point and a clustering center point, determining a primary erosion value of to-be-detected steel according to local seawater data and local environment data, and carrying out the clustering sequence division of the to-be-detected steel; comparing the primary erosion value with the historical erosion data, judging whether the primary erosion value is correct or not according to a comparison result, and re-determining the primary erosion value of the to-be-detected steel according to the erosion prediction model when the primary erosion value is judged to be incorrect; the method comprises the steps of obtaining rainfall data, conducting rainfall scoring on each piece of rainfall data according to the standard rainfall pH value, determining a compensation coefficient of a primary erosion value according to the number of the rainfall data in a rainfall erosion set, and determining the performance defect grade of to-be-detected steel based on a target erosion value. And the reliability of performance detection of the coastal steel structure is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of steel structure detection, and in particular, to a detection method and system for the performance of coastal steel structures. Background Art

[0002] With the development of infrastructure construction, steel has been widely used in the construction of ports, bridges, docks, offshore platforms, etc. in coastal areas. Due to the relatively special environmental conditions in coastal areas, steel structures are faced with the influence of erosion factors during long-term use. In severe cases, it will cause structural failure and lead to construction accidents. Steel will be affected by seawater changes, environmental precipitation, wind speed, temperature fluctuations, etc. in the coastal environment.

[0003] At present, the detection of steel structures in coastal areas mainly relies on manual experience and visual assessment. The detection results will vary due to different assessment criteria of the detection personnel. Moreover, the judgment based on manual experience in manual detection will lead to errors in the detection results, making it difficult to timely discover whether there are defects in the performance of steel structures and unable to comprehensively evaluate the potential hazards of steel structures under coastal environmental conditions.

[0004] Therefore, how to provide a detection method and system for the performance of coastal steel structures is an urgent technical problem to be solved by those skilled in the art. Summary of the Invention

[0005] In view of this, the present invention proposes a detection method and system for the performance of coastal steel structures, aiming to solve the problems that the judgment based on manual experience in manual detection will lead to errors in the detection results, making it difficult to timely discover whether there are defects in the performance of steel structures and unable to comprehensively evaluate the potential hazards of steel structures under coastal environmental conditions.

[0006] On the one hand, the present invention proposes a detection method for the performance of coastal steel structures, including:

[0007] Dividing the steel to be detected into several structural detection areas, obtaining the steel structure images of each structural detection area, obtaining the pixel points of each steel structure image, and determining the clustering center points of the steel structure images. Based on each pixel point and the clustering center points, dividing all the pixel points into clustering sequence numbers, and determining whether there are structural performance defects in the steel to be detected according to the clustering sequence number division result;

[0008] When it is determined that there are structural performance defects, obtaining the local seawater data and local environmental data, determining the primary erosion value of the steel to be detected according to the local seawater data and the local environmental data, comparing the primary erosion value with the historical erosion data, and judging whether the primary erosion value is correct according to the comparison result. When it is determined that the primary erosion value is incorrect, re-determining the primary erosion value of the steel to be detected according to the erosion prediction model;

[0009] Divide the local sea area into several precipitation collection points, obtain the precipitation data of each precipitation collection point, where the precipitation data includes precipitation acidity and precipitation amount, and determine the precipitation impact data chain of the steel to be detected based on all the precipitation data. Perform precipitation scoring on each precipitation data according to the standard precipitation acidity, and reorganize the precipitation impact data chain according to the precipitation scoring to determine the precipitation erosion data chain of the steel to be detected. Preset the erosion precipitation amount, and determine the precipitation type of each precipitation data on the precipitation erosion data chain based on the preset erosion precipitation amount;

[0010] Extract all precipitation data of the same type on the precipitation erosion data chain, construct a precipitation erosion set, determine the compensation coefficient of the primary erosion value according to the number of precipitation data in the precipitation erosion set, obtain the target erosion value, and determine the performance defect level of the steel to be detected based on the target erosion value.

[0011] Further, when dividing all pixel points into clustering number sequences based on each pixel point and the clustering center point and determining whether there are structural performance defects in the steel to be detected according to the clustering number sequence division result, it includes:

[0012] The clustering number sequence includes a first defect number sequence, a second defect number sequence, and a third defect number sequence. Determine the clustering distance between each pixel point and the clustering center point;

[0013] Divide the pixel points with a clustering distance greater than the standard clustering distance into the first defect number sequence;

[0014] Divide the pixel points with a clustering distance equal to the standard clustering distance into the second defect number sequence;

[0015] Divide the pixel points with a clustering distance less than the standard clustering distance into the third defect number sequence;

[0016] Count the first defect quantity of the pixel points in the first defect number sequence, count the second defect quantity of the pixel points in the second defect number sequence, and count the third defect quantity of the pixel points in the third defect number sequence;

[0017] Sum the first defect quantity and the third defect quantity to obtain the comprehensive defect quantity, compare the comprehensive defect quantity with the second defect quantity, and judge whether there are structural performance defects in the steel to be detected according to the comparison result;

[0018] When the comprehensive defect quantity is greater than the second defect quantity, it is determined that the steel to be detected has structural performance defects;

[0019] When the comprehensive defect quantity is less than or equal to the second defect quantity, it is determined that the steel to be detected does not have structural performance defects.

[0020] Further, when it is determined that there are structural performance defects, and local seawater data and local environmental data are obtained, when determining the primary erosion value of the steel to be detected according to the local seawater data and the environmental data, it includes:

[0021] The local seawater data includes seawater salinity and seawater pH value, and the local environmental data includes average wind speed and average temperature;

[0022] The primary erosion value is obtained from the following formula:

[0023]

[0024] Among them, E represents the primary erosion value, S represents seawater salinity, Q represents seawater pH value, V represents average wind speed, and T represents average temperature.

[0025] Further, when comparing the primary erosion value with historical erosion data and judging whether the primary erosion value is correct according to the comparison result, it includes:

[0026] Compare the primary erosion value with the historical erosion value, and judge whether the primary erosion value is correct according to the comparison result;

[0027] When the primary erosion value is greater than or equal to the historical erosion value, it is determined that the primary erosion value is correct;

[0028] When the primary erosion value is less than the historical erosion value, it is determined that the primary erosion value is incorrect.

[0029] Further, when it is determined that the primary erosion value is incorrect, re-determine the primary erosion value of the steel to be detected according to the erosion prediction model, including:

[0030] Obtain historical seawater salinity, historical seawater pH value, historical average wind speed and historical average temperature, and use the historical seawater salinity, the historical seawater pH value, the historical average wind speed and the historical average temperature as the model data set;

[0031] Divide the model data set into a model training set and a model test set, and use grid search to find the model parameters of the random forest model and establish a random forest model;

[0032] Use the model training set to fit the random forest model and perform iterative training, substitute the model test set into the random forest model after the current iterative training, and calculate the correct rate of predicting the primary erosion value;

[0033] If the accuracy rate of the random forest model after the current iterative training is less than that of the random forest model after the previous iterative training, adjust the learning rate of the random forest model after the current iterative training and continue the iterative training until the preset number of iterations is reached;

[0034] If the accuracy rate of the random forest model after the current iterative training is greater than or equal to that of the random forest model after the previous iterative training, stop the iterative training and use the random forest model after the current iterative training as the erosion prediction model;

[0035] Substitute the seawater salinity, the seawater pH value, the average wind speed, and the average temperature into the erosion prediction model to obtain the primary erosion value of the steel to be detected.

[0036] Further, when performing precipitation scoring on each precipitation data according to the standard precipitation acidity and alkalinity, and reorganizing the precipitation influence data chain according to the precipitation scoring to determine the precipitation erosion data chain of the steel to be detected, it includes:

[0037] The precipitation scoring includes an over-standard score, a compliance score, and a deviation score;

[0038] Generate the over-standard score for precipitation data with a precipitation acidity and alkalinity greater than the standard precipitation acidity and alkalinity;

[0039] Generate the compliance score for precipitation data with a precipitation acidity and alkalinity equal to the standard precipitation acidity and alkalinity;

[0040] Generate the deviation score for precipitation data with a precipitation acidity and alkalinity less than the standard precipitation acidity and alkalinity;

[0041] Delete the precipitation data with a compliance score generated on the precipitation influence data chain to obtain the precipitation erosion data chain of the steel to be detected. If there is no precipitation data with an over-standard score and a deviation score, then determine the primary erosion value as the target erosion value.

[0042] Further, when presetting the erosion precipitation amount and determining the precipitation type of each precipitation data on the precipitation erosion data chain based on the preset erosion precipitation amount, it includes:

[0043] Preset a first preset erosion precipitation amount and a second preset erosion precipitation amount, where the first preset erosion precipitation amount is greater than the second preset erosion precipitation amount;

[0044] Determine the precipitation type of precipitation data with a precipitation amount greater than or equal to the first preset erosion precipitation amount as heavy rainfall;

[0045] Determine the precipitation type of precipitation data with a precipitation amount less than the first preset erosion precipitation amount and greater than the second preset erosion precipitation amount as moderate rainfall;

[0046] Determine the precipitation type with precipitation less than or equal to the second preset erosion precipitation data as small rainfall.

[0047] Further, when extracting all precipitation data of the same type on the precipitation erosion data chain and constructing a precipitation erosion set, it includes:

[0048] The precipitation erosion set includes a first erosion set, a second erosion set, and a third erosion set;

[0049] Classify the precipitation data with the precipitation type being large rainfall and the precipitation score being the over-standard score or the deviation score into the first erosion set;

[0050] Classify the precipitation data with the precipitation type being medium rainfall and the precipitation score being the over-standard score or the deviation score into the second erosion set;

[0051] Classify the precipitation data with the precipitation type being small rainfall and the precipitation score being the over-standard score or the deviation score into the third erosion set.

[0052] Further, when determining the compensation coefficient of the primary erosion value based on the number of precipitation data in the precipitation erosion set, obtaining the target erosion value, and determining the performance defect level of the steel to be detected based on the target erosion value, it includes:

[0053] The compensation coefficient is obtained by the following formula:

[0054]

[0055] Where, F represents the compensation coefficient, N1 represents the number of precipitation data with over-standard scores in the first erosion set, N2 represents the number of precipitation data with deviation scores in the first erosion set, N3 represents the number of precipitation data with over-standard scores in the second erosion set, N4 represents the number of precipitation data with deviation scores in the second erosion set, N5 represents the number of precipitation data with over-standard scores in the third erosion set, N6 represents the number of precipitation data with deviation scores in the third erosion set, α, β, and γ represent weight coefficients, and the value ranges of α, β, and γ are (0, 1];

[0056] The target erosion value is the product value of the primary erosion value and the compensation coefficient;

[0057] Preset a first target erosion value and a second target erosion value, and the first target erosion value is greater than the second target erosion value;

[0058] When the target erosion value is greater than the first target erosion value, determine the performance defect level of the steel to be detected as a first-level defect;

[0059] When the target erosion value is less than or equal to the first target erosion value and greater than the second target erosion value, the performance defect level of the steel to be detected is determined as a secondary defect;

[0060] When the target erosion value is less than or equal to the second target erosion value, the performance defect level of the steel to be detected is determined as a tertiary defect.

[0061] Compared with the prior art, the beneficial effects of the present invention are as follows: By performing pixel clustering on the steel structure images of each structural detection area, the existing structural performance defects are accurately identified, ensuring the accuracy of the detection. Combining the local seawater data and local environmental data, the primary erosion value is obtained and verified, and the erosion prediction model is used to re-determine the primary erosion value, avoiding deviations in the detection caused by the judgment of human experience, thereby improving the reliability of the detection results. On the basis of the primary erosion value, by collecting and analyzing precipitation data and optimizing the precipitation impact data chain, a comprehensive consideration of the erosion of the steel structure is ensured, the compensation coefficient is calculated and the primary erosion value is compensated, and the relationship between precipitation data, environmental data, and seawater data is comprehensively measured, ensuring the accuracy of the target erosion value, providing a scientific and accurate basis for the maintenance, repair, and safety guarantee of the steel structure.

[0062] On the other hand, the present application also provides a detection system for the performance of coastal steel structures, which is used to apply the above-mentioned detection method for the performance of coastal steel structures, including:

[0063] A defect acquisition module, configured to divide the steel to be detected into several structural detection areas, obtain the steel structure images of each structural detection area, obtain the pixel points of each steel structure image, determine the clustering center points of the steel structure images, divide all pixel points into clustering sequence numbers based on each pixel point and the clustering center points, and determine whether there are structural performance defects in the steel to be detected according to the clustering sequence number division results;

[0064] A defect verification module, configured to, when it is determined that there are structural performance defects, obtain local seawater data and local environmental data, determine the primary erosion value of the steel to be detected according to the local seawater data and the local environmental data, compare the primary erosion value with historical erosion data, judge whether the primary erosion value is correct according to the comparison result, and when it is determined that the primary erosion value is incorrect, re-determine the primary erosion value of the steel to be detected according to the erosion prediction model;

[0065] Defect impact module, configured to divide the local sea area into several precipitation collection points, obtain precipitation data of each precipitation collection point, where the precipitation data includes precipitation pH value and precipitation amount, and determine the precipitation impact data chain of the steel to be detected according to all the precipitation data, perform precipitation scoring on each precipitation data according to the standard precipitation pH value, and reorganize the precipitation impact data chain according to the precipitation scoring to determine the precipitation erosion data chain of the steel to be detected, preset the erosion precipitation amount, and determine the precipitation type of each precipitation data on the precipitation erosion data chain based on the preset erosion precipitation amount;

[0066] Defect determination module, configured to extract all precipitation data of the same type on the precipitation erosion data chain, construct a precipitation erosion set, determine the compensation coefficient of the primary erosion value according to the number of precipitation data in the precipitation erosion set, obtain the target erosion value, and determine the performance defect level of the steel to be detected based on the target erosion value.

[0067] It can be understood that the above-mentioned method and system for detecting the performance of coastal steel structures have the same beneficial effects and will not be elaborated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] By reading the following detailed description of the preferred embodiments, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of showing the preferred embodiments and are not considered to be a limitation of the present invention. Moreover, throughout the drawings, the same reference numerals are used to represent the same components. In the drawings:

[0069] Figure 1 is a flowchart of a method for detecting the performance of coastal steel structures provided by an embodiment of the present invention;

[0070] Figure 2 is a functional block diagram of a system for detecting the performance of coastal steel structures provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0071] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the drawings. Although the exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be fully conveyed to those skilled in the art. It should be noted that, without conflict, the embodiments in the present invention and the features in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the drawings and in combination with the embodiments.

[0072] Refer to Figure 1As shown, in some embodiments of the present application, this embodiment provides a method for detecting the performance of coastal steel structures, including:

[0073] S100: Divide the steel to be detected into several structural detection areas, obtain the steel structure images of each structural detection area, obtain the pixel points of each steel structure image, determine the clustering center points of the steel structure images, divide all pixel points into clustering sequence numbers based on each pixel point and the clustering center points, and determine whether there are structural performance defects in the steel to be detected according to the clustering sequence number division results.

[0074] S200: When it is determined that there are structural performance defects, obtain the local seawater data and local environmental data, determine the primary erosion value of the steel to be detected according to the local seawater data and local environmental data, compare the primary erosion value with the historical erosion data, judge whether the primary erosion value is correct according to the comparison result, and when it is determined that the primary erosion value is incorrect, re-determine the primary erosion value of the steel to be detected according to the erosion prediction model.

[0075] S300: Divide the local sea area into several precipitation collection points, obtain the precipitation data of each precipitation collection point, the precipitation data includes precipitation pH value and precipitation amount, determine the precipitation influence data chain of the steel to be detected according to all the precipitation data, score each precipitation data according to the standard precipitation pH value, and reorganize the precipitation influence data chain according to the precipitation score to determine the precipitation erosion data chain of the steel to be detected. Preset the erosion precipitation amount, and determine the precipitation type of each precipitation data on the precipitation erosion data chain based on the preset erosion precipitation amount.

[0076] S400: Extract all precipitation data of the same type on the precipitation erosion data chain, construct a precipitation erosion set, determine the compensation coefficient of the primary erosion value according to the number of precipitation data in the precipitation erosion set, obtain the target erosion value, and determine the performance defect level of the steel to be detected based on the target erosion value.

[0077] Specifically, the description of the steel to be detected is hereinafter referred to as steel structure or the steel to be detected. First, the steel to be detected is divided into several structural detection areas. Preferably, there are 20 structural detection areas, which can be adjusted according to the actual situation. Dividing the steel to be detected can accurately analyze the structural performance. Image acquisition devices such as cameras are used to obtain the steel structure images of each structural detection area. Using image analysis algorithms, the pixel points of each steel structure image can be obtained, and thus the clustering center points of the steel structure images can be obtained. Using image clustering algorithms, all pixel points are divided into clustering sequences. The steel structure images can accurately capture the subtle erosion marks of the steel structure, and can effectively judge whether there are defects in its performance, thereby improving the detection accuracy of the steel structure. The seawater in coastal areas will erode the steel structure and cause performance defects, and the temperature and wind speed in the coastal environment will also have a certain erosive effect on the steel structure. By obtaining the local seawater data and local environmental data to determine the primary erosion value of the steel to be detected, a preliminary judgment can be made on the steel structure performance of the steel to be detected. After calculating the primary erosion value, the primary erosion value is compared with the historical erosion data to verify its rationality. If there is a certain deviation between the primary erosion value and the historical erosion data, the primary erosion value of the steel to be detected is re-determined according to the erosion prediction model, ensuring the accuracy of the steel structure detection.

[0078] It can be understood that the seawater in coastal areas itself will release some water vapor. With the action of the wind, pollutants in the atmosphere, such as sulfur dioxide and ammonia emitted from traffic in coastal cities, will combine with the water vapor of the seawater, which will cause the abnormal pH value of precipitation. The local sea area is divided into several precipitation collection points. When dividing, a relative sea area can be determined according to the environment where the steel structure is located, and this sea area is divided into several precipitation collection points. Preferably, there are 10 precipitation collection points, which can be adjusted according to the local precipitation level. According to the standard precipitation pH value, each precipitation data is scored for precipitation, and the precipitation erosion data chain of the steel to be detected is determined. By presetting the erosion precipitation, the precipitation type of each precipitation data on the precipitation erosion data chain is determined, which can comprehensively analyze different precipitation conditions and accurately evaluate the erosion impact of precipitation on the steel structure, improving the reliability and comprehensiveness of the performance detection of the steel structure. By constructing a precipitation erosion set, the compensation coefficient of the primary erosion value is determined and the target erosion value is obtained. The target erosion value combines the local environmental data, local seawater data and precipitation data, further improving the reliability and comprehensiveness of the detection. Based on the target erosion value, the performance defect level of the steel to be detected is determined, avoiding the detection errors caused by human experience, and thus comprehensively evaluating the performance of the steel structure under coastal environmental conditions.

[0079] In some embodiments of the present application, when performing clustering sequence division on all pixel points based on each pixel point and the clustering center point, and determining whether there are structural performance defects in the steel to be detected according to the clustering sequence division result, it includes: The clustering sequence includes a first defect sequence, a second defect sequence, and a third defect sequence. Determine the clustering distance between each pixel point and the clustering center point. Divide the pixel points with a clustering distance greater than the standard clustering distance into the first defect sequence, divide the pixel points with a clustering distance equal to the standard clustering distance into the second defect sequence, and divide the pixel points with a clustering distance less than the standard clustering distance into the third defect sequence. Count the first defect quantity of the pixel points in the first defect sequence, count the second defect quantity of the pixel points in the second defect sequence, and count the third defect quantity of the pixel points in the third defect sequence. Sum the first defect quantity and the third defect quantity to obtain the comprehensive defect quantity. Compare the comprehensive defect quantity with the second defect quantity, and determine whether there are structural performance defects in the steel to be detected according to the comparison result. When the comprehensive defect quantity is greater than the second defect quantity, it is determined that the steel to be detected has structural performance defects. When the comprehensive defect quantity is less than or equal to the second defect quantity, it is determined that the steel to be detected does not have structural performance defects.

[0080] Specifically, the clustering center point can be determined by one of the K-means algorithm, spectral clustering algorithm, density clustering algorithm, etc., and can be specifically selected according to the actual situation. The determination method of the clustering distance is relatively mature and lengthy, so it will not be described in detail here. The standard clustering distance represents the clustering distance between the pixel points of each steel structure image collected last time and the clustering center point. Using this as a judgment benchmark can accurately measure the differences among them, thus laying a foundation for quantifying the primary erosion value in the follow-up. The first defect sequence, the second defect sequence, and the third defect sequence represent the deviation situation between the clustering distance and the standard clustering distance, avoiding the judgment relying solely on manual experience. The calculation of the comprehensive defect quantity comprehensively considers different situations. Comparing it with the second defect quantity can reflect the proportion of the standard clustering distance, improving the accuracy of determining whether there are structural performance defects in the steel to be detected.

[0081] In some embodiments of the present application, when it is determined that there are structural performance defects, obtain the local seawater data and local environmental data, and determine the primary erosion value of the steel to be detected according to the local seawater data and environmental data, including: The local seawater data includes seawater salinity and seawater pH value, the local environmental data includes average wind speed and average temperature, and the primary erosion value is obtained from the following formula:

[0082]

[0083] Where, E represents the primary erosion value, S represents the seawater salinity, Q represents the seawater pH value, V represents the average wind speed, and T represents the average temperature.

[0084] In some embodiments of the present application, when comparing the primary erosion value with the historical erosion data and determining whether the primary erosion value is correct according to the comparison result, it includes: comparing the primary erosion value with the historical erosion value, and determining whether the primary erosion value is correct according to the comparison result. When the primary erosion value is greater than or equal to the historical erosion value, it is determined that the primary erosion value is correct; when the primary erosion value is less than the historical erosion value, it is determined that the primary erosion value is incorrect.

[0085] Specifically, the calculation of the primary erosion value comprehensively considers four erosion influencing factors: seawater salinity, seawater pH value, average wind speed, and average temperature. Higher salinity promotes the erosion process of steel structures. Moreover, the pH value of seawater has an erosive effect on steel structures. Under the influence of seawater, the wind speed and temperature in the environment promote the erosion effect. Using the average wind speed and average temperature avoids the influence of accidental situations on the primary erosion value. If the primary erosion value is greater than or equal to the historical erosion value, this indicates that under the current conditions of seawater salinity, seawater pH value, average wind speed, and average temperature, changes in environmental factors, such as higher temperature, will lead to increased erosion. At this time, the currently calculated primary erosion value is reasonable and conforms to the actual situation of environmental changes. Therefore, it can be determined that the primary erosion value is correct. If the primary erosion value is less than the historical erosion value, this situation means that the changes in the current environmental conditions have not been fully considered, or the changes in some factors have not been correctly reflected. For example, if the current seawater salinity or temperature is higher than that in the historical period, but the calculated primary erosion value is lower than the historical erosion value, this indicates that some factors may have been omitted in the calculation, or the parameters in the formula have not been updated correctly. Therefore, in this case, the primary erosion value being less than the historical erosion value indicates that there is an error in the current calculation, and it is determined that the primary erosion value is incorrect, thus ensuring the accuracy of the detection.

[0086] In some embodiments of the present application, when it is determined that the primary erosion value is incorrect, re-determining the primary erosion value of the steel to be detected according to the erosion prediction model includes: obtaining the historical seawater salinity, historical seawater pH value, historical average wind speed, and historical average temperature, using the historical seawater salinity, historical seawater pH value, historical average wind speed, and historical average temperature as the model data set, dividing the model data set into a model training set and a model test set, using grid search to find the model parameters of the random forest model and establishing the random forest model, using the model training set to fit the random forest model and performing iterative training, substituting the model test set into the random forest model after the current iterative training, calculating the correct rate of predicting the primary erosion value, if the correct rate of the random forest model after the current iterative training is less than the correct rate of the random forest model after the previous iterative training, adjusting the learning rate of the random forest model after the current iterative training, and continuing the iterative training until the preset number of iterations is reached, if the correct rate of the random forest model after the current iterative training is greater than or equal to the correct rate of the random forest model after the previous iterative training, stopping the iterative training and using the random forest model after the current iterative training as the erosion prediction model, substituting the seawater salinity, seawater pH value, average wind speed, and average temperature into the erosion prediction model to obtain the primary erosion value of the steel to be detected.

[0087] Specifically, the model data set contains local seawater data and local environmental data from different time periods. Dividing the model data set into a model training set and a model test set, with 50%-90% of the data as the model training set and the rest as the model test set, ensures that the model training set and the model test set can contain various local seawater data and local environmental data from different periods, so as to improve the generalization ability of the erosion prediction model. Grid search exhaustively searches for model parameter combinations in the parameter space, uses the model training set to fit the random forest model and performs iterative training. In each iteration, the model tries to learn the patterns and relationships in the data to improve its prediction or classification ability, thereby improving the accuracy and stability of the model prediction. Using the model test set to verify the correct rate of the random forest model after the current iterative training measures the performance of the model after iterative training.

[0088] It can be understood that if the correct rate of the random forest model after the current iterative training is less than the correct rate of the random forest model after the previous iterative training, adjusting the learning rate of the random forest model after the current iterative training helps the model to stably approach the global optimal solution, thereby improving the prediction ability. If the correct rate of the random forest model after the current iterative training is greater than or equal to the correct rate of the random forest model after the previous iterative training, it means that the performance of the model has tended to be stable, and the model after the iterative training is used as the erosion prediction model to obtain the primary erosion value of the steel to be detected, improving the reliability and accuracy of the detection.

[0089] In some embodiments of the present application, when performing precipitation scoring on each precipitation data according to the standard precipitation pH value, reorganizing the precipitation impact data chain according to the precipitation scoring, and determining the precipitation erosion data chain of the steel to be detected, it includes: The precipitation scoring includes over-standard scoring, compliance scoring, and deviation scoring. The precipitation data with a precipitation pH value greater than the standard precipitation pH value is generated with over-standard scoring, the precipitation data with a precipitation pH value equal to the standard precipitation pH value is generated with compliance scoring, the precipitation data with a precipitation pH value less than the standard precipitation pH value is generated with deviation scoring. The precipitation data with compliance scoring on the precipitation impact data chain is removed to obtain the precipitation erosion data chain of the steel to be detected. If there is no precipitation data with over-standard scoring and deviation scoring, the primary erosion value is determined as the target erosion value.

[0090] In some embodiments of the present application, when presetting the erosion precipitation amount and determining the precipitation type of each precipitation data on the precipitation erosion data chain based on the preset erosion precipitation amount, it includes: Presetting a first preset erosion precipitation amount and a second preset erosion precipitation amount, where the first preset erosion precipitation amount is greater than the second preset erosion precipitation amount. The precipitation type of the precipitation data with a precipitation amount greater than or equal to the first preset erosion precipitation amount is determined as heavy rainfall, the precipitation type of the precipitation data with a precipitation amount less than the first preset erosion precipitation amount and greater than the second preset erosion precipitation amount is determined as moderate rainfall, and the precipitation type of the precipitation data with a precipitation amount less than or equal to the second preset erosion precipitation amount is determined as light rainfall.

[0091] Specifically, each precipitation data is scored using the standard precipitation pH value, and the precipitation impact data chain is reorganized according to the precipitation scoring, thereby determining the precipitation erosion data chain of the steel to be detected. The precipitation data with compliance scoring is removed from the precipitation impact data chain, which can accurately exclude the erosion impact of the standard precipitation pH value on the steel to be detected, and recognize the erosion effect of the acidic or alkaline rainfall with a non-standard precipitation pH value on the steel to be detected, making the precipitation erosion data chain focus on the influential data. If there is no precipitation data with over-standard scoring or deviation scoring, the primary erosion value will be directly determined as the target erosion value. Moreover, by determining the precipitation type according to the relationship between the precipitation amount and the first preset erosion precipitation amount and the second preset erosion precipitation amount, the precipitation data can be further refined, thereby distinguishing the erosion impact degrees of different precipitation types on the steel structure, improving the reliability and accuracy of the steel structure detection.

[0092] In some embodiments of the present application, when extracting all precipitation data of the same type on the precipitation erosion data chain and constructing a precipitation erosion set, it includes: The precipitation erosion set includes a first erosion set, a second erosion set, and a third erosion set. Precipitation data with a precipitation type of heavy rainfall and a precipitation score of an over-standard score or a deviation score is classified into the first erosion set. Precipitation data with a precipitation type of moderate rainfall and a precipitation score of an over-standard score or a deviation score is classified into the second erosion set. Precipitation data with a precipitation type of light rainfall and a precipitation score of an over-standard score or a deviation score is classified into the third erosion set.

[0093] In some embodiments of the present application, when determining the compensation coefficient of the primary erosion value based on the number of precipitation data in the precipitation erosion set, obtaining the target erosion value, and determining the performance defect level of the steel to be detected based on the target erosion value, it includes:

[0094] The compensation coefficient is obtained by the following formula:

[0095]

[0096] Wherein, F represents the compensation coefficient, N1 represents the number of precipitation data with an over-standard score in the first erosion set, N2 represents the number of precipitation data with a deviation score in the first erosion set, N3 represents the number of precipitation data with an over-standard score in the second erosion set, N4 represents the number of precipitation data with a deviation score in the second erosion set, N5 represents the number of precipitation data with an over-standard score in the third erosion set, N6 represents the number of precipitation data with a deviation score in the third erosion set, α, β, and γ represent weight coefficients, and the value ranges of α, β, and γ are (0, 1]. The target erosion value is the product value of the primary erosion value and the compensation coefficient. A first target erosion value and a second target erosion value are preset in advance. The first target erosion value is greater than the second target erosion value. When the target erosion value is greater than the first target erosion value, the performance defect level of the steel to be detected is determined as a first-level defect. When the target erosion value is less than or equal to the first target erosion value and greater than the second target erosion value, the performance defect level of the steel to be detected is determined as a second-level defect. When the target erosion value is less than or equal to the second target erosion value, the performance defect level of the steel to be detected is determined as a third-level defect.

[0097] Specifically, the compensation coefficient is determined based on the quantities of precipitation data with different precipitation scores and different precipitation types in the first erosion set, the second erosion set, and the third erosion set. These quantities reflect the erosion influence degrees of various precipitation conditions on the steel structure. The obtained compensation coefficient can accurately measure the erosion relationship between precipitation and the steel structure, thereby compensating the primary erosion value to obtain an accurate target erosion value. By weighing the influences among precipitation, seawater, and the environment, the target erosion value enables the detection of the steel structure to conform to the current actual erosion situation, avoiding misjudgment caused by human experience. By comparing the target erosion value with the preset first target erosion value and second target erosion value, the erosion degree of the steel structure is divided into three defect levels. When the target erosion value is greater than the first target erosion value, it indicates that the erosion degree is relatively severe, and the performance of the steel structure is determined as a first-level defect. When the target erosion value is less than or equal to the first target erosion value and greater than the second target erosion value, it indicates that the erosion degree is at a medium level, and the performance of the steel structure is determined as a second-level defect. When the target erosion value is less than or equal to the second target erosion value, it indicates that the erosion degree is relatively light, and the performance of the steel structure is determined as a third-level defect. Determining the performance defect level of the steel to be detected based on the target erosion value provides a data basis for subsequent maintenance and management, avoiding the risk of serious consequences caused by the failure to timely detect the influence of erosion on the performance of the steel structure, and improving the reliability and accuracy of the detection.

[0098] In summary, the beneficial effects of the present invention are as follows: By performing pixel clustering on the steel structure images in each structural detection area, the existing structural performance defects are accurately identified, ensuring the accuracy of the detection. Combining the local seawater data and local environmental data, a primary erosion value is obtained and verified, and an erosion prediction model is used to re-determine the primary erosion value, avoiding deviation in the detection caused by the judgment of human experience, thereby improving the reliability of the detection result. Based on the primary erosion value, by collecting and analyzing precipitation data and optimizing the precipitation influence data chain, a comprehensive consideration of the erosion of the steel structure is ensured. The compensation coefficient is calculated and the primary erosion value is compensated, comprehensively weighing the relationship among precipitation data, environmental data, and seawater data, ensuring the accuracy of the target erosion value, and providing a scientific and accurate basis for the maintenance, repair, and safety guarantee of the steel structure.

[0099] In another preferred manner based on the above embodiments, refer to Figure 2 As shown, this embodiment provides a detection system for the performance of coastal steel structures, which is used to apply the above detection method for the performance of coastal steel structures, and includes:

[0100] The defect acquisition module is configured to divide the steel to be detected into several structural detection areas, acquire the steel structure images of each structural detection area, acquire the pixel points of each steel structure image, determine the clustering center points of the steel structure images, divide all the pixel points into clustering sequence numbers based on each pixel point and the clustering center points, and determine whether there are structural performance defects in the steel to be detected according to the clustering sequence number division result.

[0101] The defect verification module is configured to, when it is determined that there are structural performance defects, acquire the local seawater data and local environmental data, determine the primary erosion value of the steel to be detected according to the local seawater data and local environmental data, compare the primary erosion value with the historical erosion data, judge whether the primary erosion value is correct according to the comparison result, and when it is determined that the primary erosion value is incorrect, re-determine the primary erosion value of the steel to be detected according to the erosion prediction model.

[0102] The defect influence module is configured to divide the local sea area into several precipitation collection points, acquire the precipitation data of each precipitation collection point, the precipitation data includes precipitation pH value and precipitation amount, and determine the precipitation influence data chain of the steel to be detected according to all the precipitation data, perform precipitation scoring on each precipitation data according to the standard precipitation pH value, and reorganize the precipitation influence data chain according to the precipitation scoring to determine the precipitation erosion data chain of the steel to be detected, preset the erosion precipitation amount, and determine the precipitation type of each precipitation data on the precipitation erosion data chain based on the preset erosion precipitation amount.

[0103] The defect determination module is configured to extract all the precipitation data of the same type on the precipitation erosion data chain, construct a precipitation erosion set, determine the compensation coefficient of the primary erosion value according to the number of precipitation data in the precipitation erosion set, obtain the target erosion value, and determine the performance defect level of the steel to be detected based on the target erosion value.

[0104] Specifically, the defect acquisition module can accurately determine whether there are structural performance defects in the steel to be detected by combining image processing algorithms. By acquiring the steel structure image and clustering and dividing the pixel points in the steel structure image to determine the clustering center points, it can judge whether there are performance defects in the steel structure, improving the level of system automatic detection and reducing the dependence on human experience. The defect verification module determines the primary erosion value based on the local seawater data and local environmental data and compares it with the historical erosion data. If the primary erosion value is incorrect, it uses the erosion prediction model to re-determine it to ensure the accuracy of the detection result. The defect impact module is responsible for reorganizing the precipitation impact data chain of the steel to be detected according to the precipitation data and scoring the precipitation data through the standard precipitation pH value to determine a precipitation erosion data chain. The precipitation erosion data chain represents the precipitation situation under non-standard precipitation environments. According to the determined precipitation erosion data chain, the preset erosion precipitation is set and the precipitation type of each precipitation data on the precipitation erosion data chain is determined based on the preset erosion precipitation, which can refine different precipitation situations and measure the erosion situation of the steel structure by the precipitation pH value and precipitation amount. The defect determination module constructs a precipitation erosion set, calculates the compensation coefficient of the primary erosion value, determines the target erosion value, and based on this, obtains the performance defect level of the steel to be detected, improving the automation degree and detection accuracy of the system.

[0105] Those skilled in the art should understand that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0106] The present application is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, and the combination of the processes and / or blocks in the flowchart and / or block diagram can also be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for realizing the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0107] These computer program instructions can also be stored in a computer-readable storage device that can direct a computer or other programmable data processing apparatus to operate in a particular manner, such that the instructions stored in the computer-readable storage device produce a manufacture including an instruction device that implements the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 the functions specified in one block or multiple blocks.

[0108] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus, such that a series of operational steps are performed on the computer or other programmable apparatus to produce a computer-implemented process, so that the instructions executed on the computer or other programmable apparatus provide steps for implementing the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 the functions specified in one block or multiple blocks.

[0109] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: modifications or equivalent replacements can still be made to the specific implementation manners of the present invention, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention should be covered by the protection scope of the claims of the present invention.

Claims

1. A method for detecting the performance of coastal steel structures, characterized in that: include: The steel to be inspected is divided into several structural inspection areas, a steel structure image of each structural inspection area is obtained, and pixel points of each steel structure image are obtained, and a cluster center point of the steel structure image is determined, and all pixel points are divided into cluster series based on each pixel point and the cluster center point, and whether the steel to be inspected has structural performance defects is determined according to the cluster series division result; When it is determined that there are structural performance defects, local seawater data and local environmental data are obtained, the primary corrosion value of the steel to be tested is determined according to the local seawater data and the local environmental data, the primary corrosion value is compared with the historical corrosion data, and whether the primary corrosion value is correct is determined according to the comparison result; when it is determined that the primary corrosion value is incorrect, the primary corrosion value of the steel to be tested is re-determined according to the corrosion prediction model; Divide the local sea area into a number of precipitation collection points, obtain precipitation data of each precipitation collection point, the precipitation data includes precipitation pH and precipitation amount, and determine the precipitation impact data chain of the steel to be tested according to all precipitation data, perform precipitation scoring on each precipitation data according to the standard precipitation pH, and reorganize the precipitation impact data chain according to the precipitation score to determine the precipitation erosion data chain of the steel to be tested, preset the erosion precipitation amount, and determine the precipitation type of each precipitation data on the precipitation erosion data chain based on the preset erosion precipitation amount; Extract all precipitation data of the same type on the precipitation erosion data chain, construct a precipitation erosion set, determine the compensation coefficient of the primary erosion value according to the number of precipitation data in the precipitation erosion set, obtain the target erosion value, and determine the performance defect level of the steel to be tested based on the target erosion value.

2. The method for detecting the performance of coastal steel structures according to claim 1, characterized in that: When all the pixel points are divided into cluster series based on each pixel point and the cluster center point, and whether the steel to be inspected has structural performance defects is determined according to the cluster series division result, it includes: The cluster number sequence includes a first defect number sequence, a second defect number sequence and a third defect number sequence, and a cluster distance between each pixel point and the cluster center point is determined; Dividing the pixel points whose clustering distance is greater than the standard clustering distance into the first defect sequence; Dividing the pixel points whose clustering distance is equal to the standard clustering distance into the second defect sequence; Dividing the pixel points whose clustering distance is less than the standard clustering distance into the third defect sequence; Counting the number of first defects of pixels in the first defect sequence, counting the number of second defects of pixels in the second defect sequence, and counting the number of third defects of pixels in the third defect sequence; summing the first defect number and the third defect number to obtain a comprehensive defect number, obtaining the comprehensive defect number and comparing it with the second defect number, and judging whether the steel to be tested has structural performance defects according to the comparison result; When the comprehensive defect number is greater than the second defect number, it is determined that the steel to be tested has structural performance defects; When the comprehensive defect number is less than or equal to the second defect number, it is determined that the steel to be tested has no structural performance defects.

3. The method for detecting the performance of coastal steel structures according to claim 2, characterized in that: When it is determined that there is a structural performance defect, obtaining local seawater data and local environmental data, and determining the primary corrosion value of the steel to be tested according to the local seawater data and the environmental data, includes: The local seawater data includes seawater salinity and seawater pH value, and the local environmental data includes average wind speed and average temperature; The primary erosion value is obtained from the following formula: Among them, E represents the primary erosion value, S represents the salinity of seawater, Q represents the pH value of seawater, V represents the average wind speed, and T represents the average temperature.

4. The method for detecting the performance of coastal steel structures according to claim 3, characterized in that: When comparing the primary erosion value with the historical erosion data and judging whether the primary erosion value is correct according to the comparison result, the method includes: Comparing the primary erosion value with the historical erosion value, and judging whether the primary erosion value is correct according to the comparison result; When the primary erosion value is greater than or equal to the historical erosion value, it is determined that the primary erosion value is correct; When the primary erosion value is less than the historical erosion value, it is determined that the primary erosion value is incorrect.

5. The method for detecting the performance of coastal steel structures according to claim 4, characterized in that: When it is determined that the primary corrosion value is incorrect, the primary corrosion value of the steel to be tested is re-determined according to the corrosion prediction model, including: Acquire historical seawater salinity, historical seawater pH value, historical average wind speed and historical average temperature, and use the historical seawater salinity, the historical seawater pH value, the historical average wind speed and the historical average temperature as a model data set; The model data set is divided into a model training set and a model test set, and a grid search is used to find model parameters of a random forest model and establish a random forest model; Use the model training set to fit the random forest model and perform iterative training, substitute the model test set into the random forest model after the current iterative training, and calculate the accuracy of predicting the primary erosion value; If the accuracy of the random forest model after the current iterative training is less than the accuracy of the random forest model after the previous iterative training, the learning rate of the random forest model after the current iterative training is adjusted, and the iterative training is continued until the preset number of iterations is reached; If the accuracy of the random forest model after the current iterative training is greater than or equal to the accuracy of the random forest model after the previous iterative training, the iterative training is stopped and the random forest model after the current iterative training is used as the erosion prediction model; The seawater salinity, the seawater pH value, the average wind speed and the average temperature are substituted into the corrosion prediction model to obtain the primary corrosion value of the steel to be tested.

6. The method for detecting the performance of coastal steel structures according to claim 5, characterized in that: When each precipitation data is scored according to the standard precipitation pH, and the precipitation impact data chain is reorganized according to the precipitation score to determine the precipitation erosion data chain of the steel to be tested, it includes: The precipitation score includes an excess score, a compliance score, and a deviation score; The precipitation data with precipitation pH greater than the standard precipitation pH is used to generate the above-standard score; The compliance score is generated by using precipitation data whose precipitation pH is equal to the standard precipitation pH; Generate the deviation score for precipitation data with precipitation pH less than the standard precipitation pH; The precipitation data generated on the precipitation impact data chain that meets the score is eliminated to obtain the precipitation erosion data chain of the steel to be tested. If there is no precipitation data with the excessive score and the deviation score, the primary erosion value is determined as the target erosion value.

7. The method for detecting the performance of coastal steel structures according to claim 6, characterized in that: When the erosion precipitation is preset, and the precipitation type of each precipitation data on the precipitation erosion data chain is determined based on the preset erosion precipitation, it includes: Presetting a first preset erosion precipitation amount and a second preset erosion precipitation amount, wherein the first preset erosion precipitation amount is greater than the second preset erosion precipitation amount; Determining the precipitation type of the precipitation data whose precipitation amount is greater than or equal to the first preset erosion precipitation amount as heavy rainfall; Determine the precipitation type of precipitation data whose precipitation amount is less than the first preset erosion precipitation amount and greater than the second preset erosion precipitation amount as medium rainfall; The precipitation type of the precipitation data with a precipitation amount less than or equal to the second preset erosion precipitation amount is determined as light rainfall.

8. The method for detecting the performance of coastal steel structures according to claim 7, characterized in that: When extracting all precipitation data of the same type on the precipitation erosion data chain and constructing a precipitation erosion set, it includes: The precipitation erosion set includes a first erosion set, a second erosion set and a third erosion set; Classify the precipitation data whose precipitation type is heavy rainfall and whose precipitation score is the exceeding score or the deviation score into the first erosion set; Classify the precipitation data whose precipitation type is medium rainfall and whose precipitation score is the exceeding score or the deviation score into the second erosion set; The precipitation data whose precipitation type is light rainfall and whose precipitation score is the exceeding score or the deviation score is divided into the third erosion set.

9. The method for detecting the performance of coastal steel structures according to claim 8, characterized in that: When determining the compensation coefficient of the primary erosion value according to the amount of precipitation data in the precipitation erosion set, obtaining the target erosion value, and determining the performance defect level of the steel to be tested based on the target erosion value, the method includes: The compensation factor is obtained by the following formula: Wherein, F represents the compensation coefficient, N1 represents the number of precipitation data with over-standard scores in the first erosion set, N2 represents the number of precipitation data with deviation scores in the first erosion set, N3 represents the number of precipitation data with over-standard scores in the second erosion set, N4 represents the number of precipitation data with deviation scores in the second erosion set, N5 represents the number of precipitation data with over-standard scores in the third erosion set, N6 represents the number of precipitation data with deviation scores in the third erosion set, α, β and γ represent weight coefficients, and the value range of α, β and γ is (0, 1]; The target erosion value is the product of the primary erosion value and the compensation coefficient; Presetting a first target erosion value and a second target erosion value, wherein the first target erosion value is greater than the second target erosion value; When the target corrosion value is greater than the first target corrosion value, the performance defect level of the steel to be tested is determined as a primary defect; When the target corrosion value is less than or equal to the first target corrosion value and greater than the second target corrosion value, the performance defect level of the steel to be tested is determined as a secondary defect; When the target corrosion value is less than or equal to the second target corrosion value, the performance defect level of the steel to be tested is determined as a third-level defect.

10. A detection system for coastal steel structure performance, applied to the detection method for coastal steel structure performance as claimed in any one of claims 1 to 9, characterized in that: include: A defect acquisition module is configured to divide the steel to be inspected into a plurality of structural inspection areas, obtain a steel structure image of each structural inspection area, obtain pixel points of each steel structure image, determine a cluster center point of the steel structure image, perform cluster series division on all pixel points based on each pixel point and the cluster center point, and determine whether the steel to be inspected has structural performance defects according to the cluster series division result; a defect verification module configured to, when determining that there is a structural performance defect, obtain local seawater data and local environmental data, determine the primary corrosion value of the steel to be tested according to the local seawater data and the local environmental data, compare the primary corrosion value with historical corrosion data, and determine whether the primary corrosion value is correct according to the comparison result; and when determining that the primary corrosion value is incorrect, re-determine the primary corrosion value of the steel to be tested according to the corrosion prediction model; The defect impact module is configured to divide the local sea area into a number of precipitation collection points, obtain precipitation data of each precipitation collection point, the precipitation data includes precipitation pH and precipitation amount, and determine the precipitation impact data chain of the steel to be tested according to all precipitation data, perform precipitation scoring on each precipitation data according to the standard precipitation pH, and reorganize the precipitation impact data chain according to the precipitation score to determine the precipitation erosion data chain of the steel to be tested, preset the erosion precipitation amount, and determine the precipitation type of each precipitation data on the precipitation erosion data chain based on the preset erosion precipitation amount; The defect determination module is configured to extract all precipitation data of the same type on the precipitation erosion data chain, construct a precipitation erosion set, determine the compensation coefficient of the primary erosion value according to the number of precipitation data in the precipitation erosion set, obtain the target erosion value, and determine the performance defect level of the steel to be tested based on the target erosion value.

Images