Method and device for predicting maximum corrosion depth of storage tank bottom plate
Through a method combining generalized extreme value distribution and immunogenetic algorithm, the maximum corrosion depth of the storage tank bottom plate is calculated, which solves the problems of inaccurate prediction and large data volume in the existing technology, and achieves more efficient and accurate corrosion depth prediction.
Patent Information
- Application Number
- CN202310486143.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-05-02
- Publication Date
- 2024-11-08
AI Technical Summary
In the prior art, the prediction of the maximum corrosion depth of the storage tank base plate mainly depends on the Gumbel extreme value theory, and the probability distribution of the corrosion depth cannot be accurately described, and the parameter estimation requires a large amount of data.
A generalized extreme value distribution method based on cumulative probability is used, combined with an immune genetic algorithm to optimize the model parameters, and the target cumulative probability is calculated through the overall detection area and unit detection area, thereby determining the maximum corrosion depth of the tank bottom plate.
It improves the prediction accuracy and efficiency of the maximum corrosion depth of the storage tank base plate, solves the limitations of a single extreme distribution and large data volume, and provides higher detection rate and accuracy.
Smart Images

Figure CN118917161A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of storage tank safety, and in particular to a method and a device for predicting the maximum corrosion depth of a storage tank bottom plate. Background Art
[0002] With the advancement of science and technology and the development of the petrochemical industry, storage tanks are important energy storage equipment production facilities in the petrochemical industry. They can store a large amount of toxic, harmful, flammable, explosive and other important energy substances. At the same time, because storage tanks have the characteristics of saving investment costs, strong operability and convenient management, they play an increasingly important role in product production, transportation, storage and other processes.
[0003] In-service storage tanks inevitably have defects. They are generally subjected to cyclic loads and work in a corrosive environment, which makes them very susceptible to corrosion and corrosion defects, which in turn cause local thinning of the tank wall and reduce the effective strength of the tank. Once the tank bottom plate is corroded and perforated, resulting in oil and other substances leaking, if exposed to open flames, the leaked oil and other substances will burn and explode, further causing serious casualties and property losses. Therefore, the study of methods to predict the maximum corrosion depth of the tank bottom plate is an important support for the safe and reliable operation of the tank.
[0004] At present, the prediction research of the maximum corrosion depth of the tank bottom plate mainly uses the Gumbel extreme value theory, but the single extreme value theory cannot accurately describe the probability distribution of the corrosion depth of the tank bottom plate, and the amount of data required for parameter estimation is large. Therefore, it is urgent to provide an accurate and efficient method for predicting the maximum corrosion depth of the tank bottom plate. Summary of the invention
[0005] The invention provides a method and device for predicting the maximum corrosion depth of a tank bottom plate, which are used to solve the defects in the prior art.
[0006] The present invention provides a method for predicting the maximum corrosion depth of a tank bottom plate, comprising:
[0007] Acquire corrosion depth data of the tank bottom plate, the overall detection area of the tank bottom plate, and the unit detection area of a single plate of the tank bottom plate;
[0008] Calculating the cumulative probability corresponding to the corrosion depth data, and determining a target generalized extreme value distribution of the corrosion depth data based on the cumulative probability;
[0009] Calculating the target cumulative probability corresponding to the maximum corrosion depth of the tank bottom plate based on the overall detection area and the unit detection area;
[0010] Based on the target generalized extreme value distribution, a target corrosion depth corresponding to the target cumulative probability is determined, and the target corrosion depth is used as the maximum corrosion depth of the tank bottom plate.
[0011] According to a method for predicting the maximum corrosion depth of a tank bottom plate provided by the present invention, the target generalized extreme value distribution of the corrosion depth data is determined based on the cumulative probability, including:
[0012] Based on the generalized extreme value distribution, a statistical model for depth prediction is constructed; the independent variable of the statistical model is the corrosion depth variable, and the dependent variable of the statistical model is the cumulative probability variable;
[0013] Based on the inverse function of the statistical model, a fitness function is determined, and based on the corrosion depth data and the cumulative probability, the model parameters of the statistical model are used as antibodies, and the individual with the smallest fitness function value is used as an antigen, and an immune genetic algorithm is used to solve the target value of the model parameters;
[0014] Based on the target values of the model parameters, the target generalized extreme value distribution is determined.
[0015] According to a method for predicting the maximum corrosion depth of a tank bottom plate provided by the present invention, based on the corrosion depth data and the cumulative probability, the model parameters of the statistical model are used as antibodies, the individual with the smallest value of the fitness function is used as an antigen, and an immune genetic algorithm is used to solve the target value of the model parameter, including:
[0016] Determine an initial population, wherein the initial population includes a plurality of initial individuals;
[0017] Calculating similarities between the plurality of initial individuals, and determining an individual concentration of each initial individual based on the similarities between the plurality of initial individuals;
[0018] Based on the fitness function, an initial value of the fitness function of each initial individual is calculated, and based on the individual concentration of each initial individual and the initial value of the fitness function, an incentive degree of each initial individual is calculated;
[0019] Based on the incentive degree of each initial individual, initial individuals in the initial population are screened to obtain a first screening result, and the first screening result is subjected to a genetic crossover mutation operation to generate a new population;
[0020] Based on the immune genetic operation, a new population is continuously generated by iteration until a preset number of iterations is reached, and the antibody that minimizes the value of the fitness function is output as the target value.
[0021] According to a method for predicting the maximum corrosion depth of a tank bottom plate provided by the present invention, the fitness function is determined based on the inverse function of the statistical model, including:
[0022] Substituting the corrosion depth data into the inverse function to obtain an approximate function value corresponding to the corrosion depth data;
[0023] Calculating a root mean square error between the corrosion depth data and the approximation function value;
[0024] Based on the root mean square error, the fitness function is determined.
[0025] According to a method for predicting the maximum corrosion depth of a tank bottom plate provided by the present invention, the initial population includes a preset number of initial individuals;
[0026] Accordingly, performing a genetic crossover mutation operation on the first screening result to generate a new population includes:
[0027] Copying each individual in the first screening result into the preset number of individuals, and performing a mutation operation on the copied results based on the constraint conditions of the statistical model;
[0028] Based on the motivation degree of each individual in the variation result, the variation result is screened to obtain a second screening result;
[0029] Based on the constraint condition, randomly generate new individuals, the sum of the number of individuals in the second screening result and the number of the generated new individuals being the preset number;
[0030] Based on the second screening result and the generated new individuals, a new population is obtained.
[0031] According to a method for predicting the maximum corrosion depth of a tank bottom plate provided by the present invention, the target cumulative probability corresponding to the maximum corrosion depth of the tank bottom plate is calculated based on the overall detection area and the unit detection area, including:
[0032] Calculating a regression period of depth prediction based on the overall detection area and the unit detection area;
[0033] Based on the regression period, the target cumulative probability is calculated.
[0034] According to a method for predicting the maximum corrosion depth of a tank bottom plate provided by the present invention, the regression period of depth prediction is calculated based on the overall detection area and the unit detection area, including:
[0035] The ratio of the overall detection area to the unit detection area is calculated, and the ratio is used as the regression period.
[0036] The present invention also provides a device for predicting the maximum corrosion depth of a tank bottom plate, comprising:
[0037] An acquisition module, used for acquiring corrosion depth data of a tank bottom plate, an overall detection area of the tank bottom plate, and a unit detection area of a single plate of the tank bottom plate;
[0038] A determination module, configured to calculate a cumulative probability corresponding to the corrosion depth data, and determine a target generalized extreme value distribution of the corrosion depth data based on the cumulative probability;
[0039] A calculation module, used for calculating the target cumulative probability corresponding to the maximum corrosion depth of the tank bottom plate based on the overall detection area and the unit detection area;
[0040] The prediction module is used to determine the target corrosion depth corresponding to the target cumulative probability based on the target generalized extreme value distribution, and use the target corrosion depth as the maximum corrosion depth of the tank bottom plate.
[0041] The present invention also provides an electronic device, comprising a memory, a processor and a computer program, wherein the computer program is stored in the memory and can be run on the processor, and when the processor executes the program, the method for predicting the maximum corrosion depth of the bottom plate of a storage tank as described in any one of the above is implemented.
[0042] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the method for predicting the maximum corrosion depth of the bottom plate of a storage tank as described in any one of the above is implemented.
[0043] The present invention also provides a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the method for predicting the maximum corrosion depth of a tank bottom plate as described in any one of the above is implemented.
[0044] The present invention provides a method and device for predicting the maximum corrosion depth of a tank bottom plate. The method first obtains the corrosion depth data of the tank bottom plate, the overall detection area of the tank bottom plate, and the unit detection area of a single plate of the tank bottom plate; then calculates the cumulative probability corresponding to the corrosion depth data, and determines the target generalized extreme value distribution of the corrosion depth data based on the cumulative probability; thereafter, the target cumulative probability corresponding to the maximum corrosion depth of the tank bottom plate is calculated using the overall detection area and the unit detection area; finally, the target generalized extreme value distribution is used to determine the target corrosion depth corresponding to the target cumulative probability, and the target corrosion depth is used as the maximum corrosion depth of the tank bottom plate. The method describes the corrosion distribution of the tank bottom plate by solving the target generalized extreme value distribution, which can solve the limitations of a single extreme value distribution. By determining the target cumulative probability through the overall detection area and the unit detection area, the problems of low detection rate, poor accuracy, and large amount of data required for parameter estimation of the tank corrosion depth data can be solved, so that the accuracy of the maximum corrosion depth finally obtained is higher. BRIEF DESCRIPTION OF THE DRAWINGS
[0045] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on the drawings in the following description without paying any creative work.
[0046] Figure 1 This is one of the flow diagrams of the method for predicting the maximum corrosion depth of the bottom plate of a storage tank provided by the present invention;
[0047] Figure 2 This is the second flow diagram of the method for predicting the maximum corrosion depth of the bottom plate of a storage tank provided by the present invention;
[0048] Figure 3 It is a schematic diagram of the corresponding relationship between the fitness value and the number of iterations in the method for predicting the maximum corrosion depth of the bottom plate of a storage tank provided by the present invention, in which a new population is continuously formed through iterations of immune genetic operations;
[0049] Figure 4 It is a structural schematic diagram of a device for predicting the maximum corrosion depth of a tank bottom plate provided by the present invention;
[0050] Figure 5 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION
[0051] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0052] The existing research on the prediction of the maximum corrosion depth of the tank bottom plate mainly uses the Gumbel extreme value theory, but the single extreme value theory cannot accurately describe the probability distribution of the corrosion depth of the tank bottom plate, and the amount of data required for parameter estimation is large. Based on this, an accurate and efficient method for predicting the maximum corrosion depth of the tank bottom plate is provided in an embodiment of the present invention.
[0053] Figure 1 A method for predicting the maximum corrosion depth of a tank bottom plate provided in an embodiment of the present invention includes:
[0054] S1, obtaining corrosion depth data of the tank bottom plate, the overall detection area of the tank bottom plate, and the unit detection area of a single plate of the tank bottom plate;
[0055] S2, calculating the cumulative probability corresponding to the corrosion depth data, and determining the target generalized extreme value distribution of the corrosion depth data based on the cumulative probability;
[0056] S3, calculating the target cumulative probability corresponding to the maximum corrosion depth of the tank bottom plate based on the overall detection area and the unit detection area;
[0057] S4, based on the target generalized extreme value distribution, determining the target corrosion depth corresponding to the target cumulative probability, and using the target corrosion depth as the maximum corrosion depth of the tank bottom plate.
[0058] Specifically, the method for predicting the maximum corrosion depth of the tank bottom provided in the embodiment of the present invention is executed by a device for predicting the maximum corrosion depth of the tank bottom. The device can be configured in a computer, which can be a local computer or a cloud computer. The local computer can be a computer, a tablet, etc., which is not specifically limited here.
[0059] First, step S1 is performed to obtain the corrosion depth data of the tank bottom plate, the overall detection area of the tank bottom plate, and the unit detection area of a single plate of the tank bottom plate. Here, the storage tank can be a crude oil storage tank or a storage tank for storing other resources, which is not specifically limited here.
[0060] The corrosion depth data is the detection result of the corrosion depth of the tank bottom plate, which may include multiple data. After the corrosion depth data of the tank bottom plate is detected, the bottom plate corrosion depth data are sorted from large to small, and the first n corrosion depth data are selected as sample data. Multiple corrosion depth data can be arranged in order from large to small to obtain a data sequence, which can be expressed as Y = {y1, y2, ..., y i ,...,y n}, 1≤i≤n, y i is the i-th corrosion depth data, i is y i The ordinal number in the data sequence, n is the number of corrosion depth data. In order to ensure the subsequent accurate determination of the target generalized extreme value distribution, the number of corrosion depth data n can be greater than or equal to 30. Each corrosion depth data can include the corrosion depth value, in addition to the corrosion location. The bottom plate of the tank is usually large in area and is welded by multiple middle plates and edge plates. The shape of the tank bottom plate is usually circular, and a single middle plate or edge plate is usually a quadrilateral.
[0061] In addition to obtaining the corrosion depth data of the tank bottom plate, the overall detection area of the tank bottom plate and the unit detection area of a single plate can also be obtained. The overall detection area is the total area of the tank bottom plate, and the unit detection area is the area of a single plate. A single plate can be a middle plate or an edge plate, which is not specifically limited here.
[0062] Then, step S2 is executed to calculate the cumulative probability corresponding to the corrosion depth data. The n sample data selected in step S1 are sorted from small to large, and the updated data sequence can be expressed as X = {x1, x2, ..., x i ,...,x n}, and calculate the cumulative probability. Here, each corrosion depth data corresponds to a cumulative probability, and the calculation of the cumulative probability can be realized by combining the ordinal number of each corrosion depth data in the data sequence and the number of corrosion depth data. The calculation formula can be:
[0063]
[0064] Among them, GEV(x i ) is the cumulative probability corresponding to the i-th corrosion depth data.
[0065] Since the corrosion depth distribution of the tank bottom plate is complex, and a single extreme value distribution is difficult to describe the corrosion depth distribution of the tank bottom plate and is inconsistent with the actual situation, the generalized extreme value distribution is introduced in the embodiment of the present invention to describe the corrosion distribution of the tank bottom plate, which can solve the limitations of the single extreme value distribution.
[0066] In the embodiment of the present invention, the generalized extreme value distribution can be solved by using the cumulative probability corresponding to each corrosion depth data to obtain the target generalized extreme value distribution of the corrosion depth data, and the target generalized extreme value distribution can be used to characterize the corresponding relationship between each corrosion depth data and its corresponding cumulative probability. Among them, the generalized extreme value distribution can be expressed as:
[0067]
[0068] The constraints of the generalized extreme value distribution can be expressed as:
[0069]
[0070] Among them, x is the corrosion depth variable, GEV(x) is the cumulative probability variable, k is the threshold parameter, μ is the location parameter, and σ is the scale parameter. k, μ and σ are all parameters that need to be solved.
[0071] The target values of k, μ and σ are obtained by using the cumulative probability corresponding to each corrosion depth data and then inserted into formula (2) to obtain the target generalized extreme value distribution.
[0072] Thereafter, step S3 is executed to calculate the target cumulative probability corresponding to the maximum corrosion depth of the tank bottom plate using the overall detection area of the tank bottom plate and the unit detection area of a single plate. Here, the target cumulative probability can be determined based on the inverse of the ratio of the overall detection area to the unit detection area. By using the overall detection area and the unit detection area, the target cumulative probability can be determined by estimating a large sample with a small sample, which can solve the problems of low detection rate, poor accuracy, and large amount of data required for parameter estimation of tank corrosion depth data.
[0073] Finally, step S4 is performed to determine the target corrosion depth corresponding to the target cumulative probability using the target generalized extreme value distribution. That is, the target cumulative probability is substituted into the target generalized extreme value distribution, and the corrosion depth data obtained is the target corrosion depth. The target corrosion depth can be used as the maximum corrosion depth of the tank bottom plate.
[0074] The maximum corrosion depth prediction method of the tank bottom provided in the embodiment of the present invention first obtains the corrosion depth data of the tank bottom, the overall detection area of the tank bottom and the unit detection area of a single plate of the tank bottom; then calculates the cumulative probability corresponding to the corrosion depth data, and determines the target generalized extreme value distribution of the corrosion depth data based on the cumulative probability; thereafter, the target cumulative probability corresponding to the maximum corrosion depth of the tank bottom is calculated using the overall detection area and the unit detection area; finally, the target generalized extreme value distribution is used to determine the target corrosion depth corresponding to the target cumulative probability, and the target corrosion depth is used as the maximum corrosion depth of the tank bottom. This method describes the corrosion distribution of the tank bottom by solving the target generalized extreme value distribution, which can solve the limitations of a single extreme value distribution. By determining the target cumulative probability through the overall detection area and the unit detection area, the problems of low detection rate, poor accuracy and large amount of data required for parameter estimation of the tank corrosion depth data can be solved, so that the accuracy of the maximum corrosion depth finally obtained is higher.
[0075] On the basis of the above-mentioned embodiment, the maximum corrosion depth prediction method of the bottom plate of a storage tank provided in the embodiment of the present invention, the target generalized extreme value distribution of the corrosion depth data is determined based on the cumulative probability, including:
[0076] Based on the generalized extreme value distribution, a statistical model for depth prediction is constructed; the independent variable of the statistical model is the corrosion depth data, and the dependent variable of the statistical model is the cumulative probability;
[0077] Based on the inverse function of the statistical model, a fitness function is determined, and based on the corrosion depth data and the cumulative probability, the model parameters of the statistical model are used as antibodies, and the individual with the smallest fitness function value is used as an antigen, and an immune genetic algorithm is used to solve the target value of the model parameters;
[0078] Based on the target values of the model parameters, the target generalized extreme value distribution is determined.
[0079] Specifically, when determining the target generalized extreme value distribution of corrosion depth data, the generalized extreme value distribution can be used to construct a statistical model for depth prediction. Here, the generalized extreme value distribution can be directly used as the statistical model for depth prediction, as shown in formula (2). The independent variable of the statistical model is the corrosion depth variable, that is, x, and the dependent variable of the statistical model is the cumulative probability variable, that is, GEV(x). The model parameters of the statistical model include k, μ, and σ.
[0080] The inverse function of the statistical model can be expressed as:
[0081]
[0082] After that, the fitness function is determined using the inverse function of the statistical model. Here, the independent variable of the inverse function is the corrosion depth variable, and the dependent variable is the approximate function variable of the corrosion depth variable. Substituting each corrosion depth data as the value of the corrosion depth variable into the inverse function, the corresponding approximate function value can be obtained, and then the fitness function can be determined using each corrosion depth data and its approximate function value.
[0083] The fitness function can be characterized by the error between each corrosion depth data and its approximate function value, and the error can be calculated by any one of the residual sum of squares (RSS), mean square error (MSE), root mean square error (RMSE), mean absolute error (MAE) and standard deviation (SD).
[0084] After that, the model parameters of the statistical model are used as antibodies, and the individuals with the smallest fitness function value are used as antigens, and the immune genetic algorithm is used to solve the target value of the model parameters. Among them, the immune genetic algorithm (Immune Genetic Algorithm, IGA) is a multidisciplinary cross-cutting and infiltrating optimization algorithm that combines the advantages of immune theory (Immune Algorithm, IA) and basic genetic algorithm (Simple Genetic Algorithm, SGA). This immune genetic algorithm not only retains the search characteristics of the genetic algorithm, but also utilizes the adaptive characteristics of the immune algorithm's multi-mechanism to solve the optimal solution of multiple objective functions, which can improve the fitness of the population and the convergence speed of the algorithm, and solve the problems of slow search process, reduced population diversity, and falling into local optimal solutions in the genetic algorithm.
[0085] Finally, the target values of the model parameters are used to determine the target generalized extreme value distribution. That is, the target values of the model parameters are substituted into formula (2) to obtain the target generalized extreme value distribution.
[0086] In the embodiment of the present invention, the immune genetic algorithm is used to quickly solve the target values of the model parameters, thereby improving the prediction efficiency and prediction accuracy of the maximum corrosion depth of the tank bottom plate.
[0087] On the basis of the above-mentioned embodiment, the maximum corrosion depth prediction method of the bottom plate of a storage tank provided in the embodiment of the present invention, based on the corrosion depth data and the cumulative probability, takes the model parameters of the statistical model as antibodies, takes the individual with the smallest value of the fitness function as antigen, and uses the immune genetic algorithm to solve the target value of the model parameters, including:
[0088] Determine an initial population, wherein the initial population includes a plurality of initial individuals;
[0089] Calculating similarities between the plurality of initial individuals, and determining an individual concentration of each initial individual based on the similarities between the plurality of initial individuals;
[0090] Based on the fitness function, an initial value of the fitness function of each initial individual is calculated, and based on the individual concentration of each initial individual and the initial value of the fitness function, an incentive degree of each initial individual is calculated;
[0091] Based on the incentive degree of each initial individual, initial individuals in the initial population are screened to obtain a first screening result, and the first screening result is subjected to a genetic crossover mutation operation to generate a new population;
[0092] Based on the immune genetic operation, a new population is continuously generated by iteration until a preset number of iterations is reached, and the antibody that minimizes the value of the fitness function is output as the target value.
[0093] Specifically, in the embodiment of the present invention, when the immune genetic algorithm is used to solve the target value of the model parameter, the population size N, similarity threshold M, affinity coefficient α, concentration coefficient β and preset number of iterations E of the immune genetic algorithm are first set.
[0094] Then, an initial population is determined, which may include multiple initial individuals. The number of initial individuals in the initial population may be equal to the preset population number N. The initial individuals may serve as antibodies that can bind to the antigen. The initial population A may be expressed as:
[0095] A={A1,A2,...,A N}={(k1,μ1,σ1),(k2,μ2,σ2),...,(k N ,μ N ,σ N )}; (5)
[0096] Among them, A1, A2, ..., AN are the 1st, 2nd, ..., Nth initial individuals in the initial population A respectively.
[0097] After that, the similarity between multiple initial individuals is calculated, where the similarity between any two initial individuals can be represented by the distance between any two initial individuals. Furthermore, the similarity between any two initial individuals a and b can be calculated by the following formula:
[0098]
[0099] Among them, the initial individual a is (k a ,μa ,σ a ), the initial individual b is (k b ,μ b ,σ b ).
[0100] Thereafter, the similarity between any two initial individuals is compared with the similarity threshold, and the individual concentration of any initial individual is determined based on the comparison result.
[0101] If the result of the comparison is that the similarity is less than the similarity threshold M, the similarity between any two initial individuals is reassigned to 1. If the result of the comparison is that the similarity is greater than or equal to the similarity threshold M, the similarity between any two initial individuals is reassigned to 0, that is:
[0102]
[0103] in, is the similarity between any two initial individuals a and b after reassignment.
[0104] After that, the similarity between any initial individual and other initial individuals after reassignment is added and divided by the number of initial individuals in the initial population to obtain the individual concentration of any initial individual. For example, the initial individual a is used as any initial individual, and the individual concentration ρa of the initial individual a can be calculated by the following formula:
[0105]
[0106] After that, the fitness function is used to calculate the initial value of the fitness function of each initial individual, and each initial individual is substituted into the fitness function to obtain the initial value of the fitness function of each initial individual. Furthermore, since the individual with the smallest fitness function value is the antigen, the incentive degree of each initial individual can be calculated using the individual concentration of each initial individual and the initial value of the fitness function. Taking the initial individual a as any initial individual, the calculation formula of the incentive degree Ia of the initial individual a can be expressed as:
[0107] I a = -α*RMSE a -β*ρ a ,a=1,2,...,N; (9)
[0108] Afterwards, the initial individuals in the initial population are screened using the incentive degree of each initial individual to obtain a first screening result. Here, the initial individuals in the initial population can be sorted in descending order according to the incentive degree, and then the first 50% of the initial individuals with the largest incentive degree in the initial population are selected to form the first screening result.
[0109] Afterwards, the first screening result is subjected to a genetic crossover mutation operation to generate a new population. Here, each individual in the first screening result can generate a new population through the genetic crossover mutation operation.
[0110] Afterwards, the immune genetic operation is used to continuously iterate the above process to generate a new population until the preset number of iterations is reached, and the antibody that minimizes the value of the fitness function is output as the target value.
[0111] In an embodiment of the present invention, the individual concentration is calculated to determine the individual incentive degree, and then the genetic crossover mutation operation is performed on the individual screening, so that the evolution of the population in the immune genetic algorithm no longer relies solely on genetic crossover mutation, but adds the calculation of the antibody concentration and the initial value of the fitness function to maintain the diversity of the population, thereby making the target values of the model parameters obtained by solving more accurate and reliable.
[0112] On the basis of the above-mentioned embodiment, the method for predicting the maximum corrosion depth of the bottom plate of a storage tank provided in the embodiment of the present invention, wherein the fitness function is determined based on the inverse function of the statistical model, comprises:
[0113] Substituting the corrosion depth data into the inverse function to obtain an approximate function value corresponding to the corrosion depth data;
[0114] Calculating a root mean square error between the corrosion depth data and the approximation function value;
[0115] Based on the root mean square error, the fitness function is determined.
[0116] Specifically, in the embodiment of the present invention, when determining the fitness function, the root mean square error between the corrosion depth data and the approximation function value may be calculated first, and the calculation formula may be:
[0117]
[0118] Thereafter, the calculated root mean square error can be used as a fitness function.
[0119] In the embodiment of the present invention, the root mean square error is used as the fitness function, and the errors between all corrosion depth data and the approximation function value can be systematically considered, so that the fitness function is more suitable for measuring the adaptability of individuals in the immune genetic algorithm.
[0120] On the basis of the above-mentioned embodiment, in the method for predicting the maximum corrosion depth of the bottom plate of a storage tank provided in the embodiment of the present invention, the initial population includes a preset number of the initial individuals;
[0121] Accordingly, performing a genetic crossover mutation operation on the first screening result to generate a new population includes:
[0122] Copying each individual in the first screening result into the preset number of individuals, and performing a mutation operation on the copied results based on the constraint conditions of the statistical model;
[0123] Based on the motivation degree of each individual in the variation result, the variation result is screened to obtain a second screening result;
[0124] Based on the constraint condition, randomly generate new individuals, the sum of the number of individuals in the second screening result and the number of the generated new individuals being the preset number;
[0125] Based on the second screening result and the generated new individuals, a new population is obtained.
[0126] Specifically, the initial population includes a preset number of initial individuals. When the first screening result is subjected to a genetic crossover mutation operation to generate a new population, each individual in the first screening result can be firstly copied into a preset number of individuals, i.e., N individuals. Then, using the constraint conditions of the statistical model, a mutation operation is performed on the copied result, and the mutation result also includes N individuals. The constraint conditions of the statistical model are formula (3), and the mutation operation can include single-point mutation and multi-point mutation.
[0127] After that, the results obtained by mutation are screened by using the motivation degree of each individual in the results obtained by mutation to obtain the second screening results. The screening here also first arranges the individuals in the results obtained by mutation in descending order according to the motivation degree, and selects the first 50% of individuals with high motivation degree.
[0128] Thereafter, new individuals are randomly generated using the constraints of the statistical model. That is, the randomly generated new individuals need to meet the constraints of the statistical model. The sum of the number of the generated new individuals and the number of individuals in the second screening result is a preset number, and the preset number of individuals constitutes a new population.
[0129] In the embodiment of the present invention, the genetic crossover mutation operation is used to ensure the diversity of the population, thereby making the target values of the solved model parameters more accurate and reliable.
[0130] On the basis of the above-mentioned embodiment, the maximum corrosion depth prediction method of the tank bottom plate provided in the embodiment of the present invention, the target cumulative probability corresponding to the maximum corrosion depth of the tank bottom plate is calculated based on the overall detection area and the unit detection area, including:
[0131] Calculating a regression period of depth prediction based on the overall detection area and the unit detection area;
[0132] Based on the regression period, the target cumulative probability is calculated.
[0133] Specifically, when calculating the target cumulative probability, the regression period of depth prediction can be calculated by first using the overall detection area and the unit detection area. The regression period can be the ratio of the overall detection area to the unit detection area, that is,
[0134]
[0135] Among them, T(y) is the regression period, S is the overall detection area, and s is the unit detection area.
[0136] Finally, the target cumulative probability is calculated using the regression cycle. The calculation formula for the target cumulative probability can be:
[0137]
[0138] Finally, the maximum corrosion depth of the tank bottom is x max It can be calculated by the following formula:
[0139]
[0140] like Figure 2 FIG. 1 is a schematic diagram of a complete process of a method for predicting the maximum corrosion depth of a tank bottom plate provided in an embodiment of the present invention. The method includes:
[0141] Acquire inspection data of the tank bottom plate, the inspection data including corrosion depth data, the overall inspection area of the tank bottom plate, and the unit inspection area of a single plate of the tank bottom plate;
[0142] Calculate the cumulative probability corresponding to the corrosion depth data;
[0143] Construct a statistical model for depth prediction and determine the inverse function of the statistical model;
[0144] Determine the fitness function based on the inverse function of the statistical model;
[0145] Determine the initial population;
[0146] Calculate individual concentrations;
[0147] Calculate the degree of motivation;
[0148] Perform genetic crossover and mutation operations;
[0149] Determine target values for model parameters;
[0150] Calculate the regression period of the depth prediction;
[0151] Calculate the target cumulative probability corresponding to the maximum corrosion depth;
[0152] Determine the maximum corrosion depth of the tank floor.
[0153] Taking the selected 30 corrosion depth data as an example, the cumulative probability corresponding to each corrosion depth data is calculated, as shown in Table 1.
[0154] Table 1 Corrosion depth data and its cumulative probability
[0155]
[0156] The population size N of the immune genetic algorithm is set to 20, the similarity threshold M to 0.0009, the affinity coefficient α to 2, the concentration coefficient β to 1, and the preset number of iterations E to 100.
[0157] Through the continuous iteration of immune genetic operations, a new population is formed. The corresponding relationship between the fitness value and the number of iterations is as follows: Figure 3 As shown, after about 10 iterations, the stable value is reached, and the convergence speed is very fast. When the preset number of iterations E = 100 is reached, the antibody that minimizes the fitness function is output, that is, the threshold parameter k = -0.2537, the location parameter μ = 3.5104, and the scale parameter σ = 1.5421. Furthermore, the target generalized extreme value distribution can be expressed as:
[0158]
[0159] The regression cycle is calculated based on the overall detection area of the tank bottom plate and the unit detection area of a single plate:
[0160]
[0161] According to the regression period T(y), calculate the maximum corrosion depth x max The corresponding target cumulative probability is:
[0162]
[0163] Finally, the maximum corrosion depth x max Substitute the corresponding target cumulative probability into formula (14) to calculate the maximum corrosion depth x of the tank bottom plate max for:
[0164]
[0165] When 30 data points are taken as sample data, the maximum corrosion depth calculated is 8.1381; when 20 data points are taken as sample data, the maximum corrosion depth calculated is 8.1702, and the error between the two is 0.39%. It can be seen that the prediction accuracy of the present invention is better when 10 sample data are missing.
[0166] In summary, the method for predicting the maximum corrosion depth of the tank bottom provided in the embodiment of the present invention can not only solve the limitations brought by the single extreme value distribution of the corrosion depth, but also use a small sample to estimate a large sample. At the same time, the adopted immune genetic algorithm has a fast convergence speed and can quickly predict the maximum corrosion depth of the crude oil tank bottom, thereby providing guidance for the prevention and control of corrosion and leakage of the tank bottom.
[0167] like Figure 4 As shown, based on the above embodiment, an embodiment of the present invention provides a device for predicting the maximum corrosion depth of a tank bottom plate, comprising:
[0168] An acquisition module 41 is used to acquire corrosion depth data of a tank bottom plate, an overall detection area of the tank bottom plate, and a unit detection area of a single plate of the tank bottom plate;
[0169] A determination module 42, configured to calculate a cumulative probability corresponding to the corrosion depth data, and determine a target generalized extreme value distribution of the corrosion depth data based on the cumulative probability;
[0170] A calculation module 43, configured to calculate a target cumulative probability corresponding to a maximum corrosion depth of the tank bottom plate based on the overall detection area and the unit detection area;
[0171] The prediction module 44 is used to determine the target corrosion depth corresponding to the target cumulative probability based on the target generalized extreme value distribution, and use the target corrosion depth as the maximum corrosion depth of the tank bottom plate.
[0172] On the basis of the above-mentioned embodiment, in the device for predicting the maximum corrosion depth of the bottom plate of a storage tank provided in the embodiment of the present invention, the determination module is specifically used for:
[0173] Based on the generalized extreme value distribution, a statistical model for depth prediction is constructed; the independent variable of the statistical model is the corrosion depth variable, and the dependent variable of the statistical model is the cumulative probability variable;
[0174] Based on the inverse function of the statistical model, a fitness function is determined, and based on the corrosion depth data and the cumulative probability, the model parameters of the statistical model are used as antibodies, and the individual with the smallest fitness function value is used as an antigen, and an immune genetic algorithm is used to solve the target value of the model parameters;
[0175] Based on the target values of the model parameters, the target generalized extreme value distribution is determined.
[0176] On the basis of the above-mentioned embodiment, in the device for predicting the maximum corrosion depth of the bottom plate of a storage tank provided in the embodiment of the present invention, the determination module is further specifically used for:
[0177] Determine an initial population, wherein the initial population includes a plurality of initial individuals;
[0178] Calculating similarities between the plurality of initial individuals, and determining an individual concentration of each initial individual based on the similarities between the plurality of initial individuals;
[0179] Based on the fitness function, an initial value of the fitness function of each initial individual is calculated, and based on the individual concentration of each initial individual and the initial value of the fitness function, an incentive degree of each initial individual is calculated;
[0180] Based on the incentive degree of each initial individual, initial individuals in the initial population are screened to obtain a first screening result, and the first screening result is subjected to a genetic crossover mutation operation to generate a new population;
[0181] Based on the immune genetic operation, a new population is continuously generated by iteration until a preset number of iterations is reached, and the antibody that minimizes the value of the fitness function is output as the target value.
[0182] On the basis of the above-mentioned embodiment, in the device for predicting the maximum corrosion depth of the bottom plate of a storage tank provided in the embodiment of the present invention, the determination module is further specifically used for:
[0183] Substituting the corrosion depth data into the inverse function to obtain an approximate function value corresponding to the corrosion depth data;
[0184] Calculating a root mean square error between the corrosion depth data and the approximation function value;
[0185] Based on the root mean square error, the fitness function is determined.
[0186] On the basis of the above-mentioned embodiment, in the device for predicting the maximum corrosion depth of the bottom plate of a storage tank provided in the embodiment of the present invention, the initial population includes a preset number of the initial individuals;
[0187] Accordingly, the determining module is further specifically configured to:
[0188] Copying each individual in the first screening result into the preset number of individuals, and performing a mutation operation on the copied results based on the constraint conditions of the statistical model;
[0189] Based on the motivation degree of each individual in the variation result, the variation result is screened to obtain a second screening result;
[0190] Based on the constraint condition, randomly generate new individuals, the sum of the number of individuals in the second screening result and the number of the generated new individuals being the preset number;
[0191] Based on the second screening result and the generated new individuals, a new population is obtained.
[0192] On the basis of the above-mentioned embodiment, in the device for predicting the maximum corrosion depth of the bottom plate of a storage tank provided in the embodiment of the present invention, the calculation module is specifically used for:
[0193] Calculating a regression period of depth prediction based on the overall detection area and the unit detection area;
[0194] Based on the regression period, the target cumulative probability is calculated.
[0195] On the basis of the above-mentioned embodiment, in the device for predicting the maximum corrosion depth of the bottom plate of a storage tank provided in the embodiment of the present invention, the calculation module is further specifically used for:
[0196] The ratio of the overall detection area to the unit detection area is calculated, and the ratio is used as the regression period.
[0197] Specifically, the functions of each module in the device for predicting the maximum corrosion depth of the bottom plate of a storage tank provided in the embodiment of the present invention correspond one-to-one to the operating procedures of each step in the above-mentioned method embodiment, and the effects achieved are also consistent. Please refer to the above-mentioned embodiment for details, which will not be described in detail in the embodiment of the present invention.
[0198] Figure 5 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 5 As shown, the electronic device may include: a processor (Processor) 510, a communication interface (Communications Interface) 520, a memory (Memory) 530 and a communication bus 540, wherein the processor 510, the communication interface 520, and the memory 530 complete mutual communication through the communication bus 540. The processor 510 can call the logic instructions in the memory 530 to execute the maximum corrosion depth prediction method of the tank bottom provided in the above embodiments, the method comprising: obtaining the corrosion depth data of the tank bottom, the overall detection area of the tank bottom, and the unit detection area of a single plate of the tank bottom; calculating the cumulative probability corresponding to the corrosion depth data, and determining the target generalized extreme value distribution of the corrosion depth data based on the cumulative probability; calculating the target cumulative probability corresponding to the maximum corrosion depth of the tank bottom based on the overall detection area and the unit detection area; determining the target corrosion depth corresponding to the target cumulative probability based on the target generalized extreme value distribution, and taking the target corrosion depth as the maximum corrosion depth of the tank bottom.
[0199] In addition, the logic instructions in the above-mentioned memory 530 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or the part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program codes.
[0200] On the other hand, the present invention also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the maximum corrosion depth prediction method for the tank bottom plate provided in the above-mentioned embodiments, and the method includes: obtaining the corrosion depth data of the tank bottom plate, the overall detection area of the tank bottom plate, and the unit detection area of a single plate of the tank bottom plate; calculating the cumulative probability corresponding to the corrosion depth data, and determining the target generalized extreme value distribution of the corrosion depth data based on the cumulative probability; calculating the target cumulative probability corresponding to the maximum corrosion depth of the tank bottom plate based on the overall detection area and the unit detection area; determining the target corrosion depth corresponding to the target cumulative probability based on the target generalized extreme value distribution, and using the target corrosion depth as the maximum corrosion depth of the tank bottom plate.
[0201] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which is implemented when the computer program is executed by a processor to execute the maximum corrosion depth prediction method for the tank bottom plate provided in the above-mentioned embodiments, the method comprising: obtaining the corrosion depth data of the tank bottom plate, the overall detection area of the tank bottom plate, and the unit detection area of a single plate of the tank bottom plate; calculating the cumulative probability corresponding to the corrosion depth data, and determining the target generalized extreme value distribution of the corrosion depth data based on the cumulative probability; calculating the target cumulative probability corresponding to the maximum corrosion depth of the tank bottom plate based on the overall detection area and the unit detection area; determining the target corrosion depth corresponding to the target cumulative probability based on the target generalized extreme value distribution, and using the target corrosion depth as the maximum corrosion depth of the tank bottom plate.
[0202] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.
[0203] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.
[0204] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A method for predicting the maximum corrosion depth of a tank bottom plate, characterized in that: include: Acquire corrosion depth data of the tank bottom plate, the overall detection area of the tank bottom plate, and the unit detection area of a single plate of the tank bottom plate; Calculating the cumulative probability corresponding to the corrosion depth data, and determining a target generalized extreme value distribution of the corrosion depth data based on the cumulative probability; Calculating the target cumulative probability corresponding to the maximum corrosion depth of the tank bottom plate based on the overall detection area and the unit detection area; Based on the target generalized extreme value distribution, a target corrosion depth corresponding to the target cumulative probability is determined, and the target corrosion depth is used as the maximum corrosion depth of the tank bottom plate.
2. The method for predicting the maximum corrosion depth of the tank bottom plate according to claim 1 is characterized in that: Determining a target generalized extreme value distribution of the corrosion depth data based on the cumulative probability includes: Based on the generalized extreme value distribution, a statistical model for depth prediction is constructed; the independent variable of the statistical model is the corrosion depth variable, and the dependent variable of the statistical model is the cumulative probability variable; Based on the inverse function of the statistical model, a fitness function is determined, and based on the corrosion depth data and the cumulative probability, the model parameters of the statistical model are used as antibodies, and the individual with the smallest fitness function value is used as an antigen, and an immune genetic algorithm is used to solve the target value of the model parameters; Based on the target values of the model parameters, the target generalized extreme value distribution is determined.
3. The method for predicting the maximum corrosion depth of the tank bottom plate according to claim 2 is characterized in that: The method of solving the target value of the model parameter based on the corrosion depth data and the cumulative probability, taking the model parameter of the statistical model as an antibody, taking the individual with the smallest value of the fitness function as an antigen, and using an immune genetic algorithm to solve the target value of the model parameter includes: Determine an initial population, wherein the initial population includes a plurality of initial individuals; Calculating similarities between the plurality of initial individuals, and determining an individual concentration of each initial individual based on the similarities between the plurality of initial individuals; Based on the fitness function, an initial value of the fitness function of each initial individual is calculated, and based on the individual concentration of each initial individual and the initial value of the fitness function, an incentive degree of each initial individual is calculated; Based on the incentive degree of each initial individual, initial individuals in the initial population are screened to obtain a first screening result, and the first screening result is subjected to a genetic crossover mutation operation to generate a new population; Based on the immune genetic operation, a new population is continuously generated by iteration until a preset number of iterations is reached, and the antibody that minimizes the value of the fitness function is output as the target value.
4. The method for predicting the maximum corrosion depth of the tank bottom plate according to claim 3 is characterized in that: Determining the fitness function based on the inverse function of the statistical model includes: Substituting the corrosion depth data into the inverse function to obtain an approximate function value corresponding to the corrosion depth data; Calculating a root mean square error between the corrosion depth data and the approximation function value; Based on the root mean square error, the fitness function is determined.
5. The method for predicting the maximum corrosion depth of the tank bottom plate according to claim 3 is characterized in that: The initial population includes a preset number of initial individuals; Accordingly, performing a genetic crossover mutation operation on the first screening result to generate a new population includes: Copying each individual in the first screening result into the preset number of individuals, and performing a mutation operation on the copied results based on the constraint conditions of the statistical model; Based on the motivation degree of each individual in the variation result, the variation result is screened to obtain a second screening result; Based on the constraint condition, randomly generate new individuals, the sum of the number of individuals in the second screening result and the number of the generated new individuals being the preset number; Based on the second screening result and the generated new individuals, a new population is obtained.
6. The method for predicting the maximum corrosion depth of a tank bottom plate according to any one of claims 1 to 5, characterized in that: The calculating, based on the overall detection area and the unit detection area, the target cumulative probability corresponding to the maximum corrosion depth of the tank bottom plate comprises: Calculating a regression period of depth prediction based on the overall detection area and the unit detection area; Based on the regression period, the target cumulative probability is calculated.
7. The method for predicting the maximum corrosion depth of the tank bottom plate according to claim 6 is characterized in that: The calculating the regression period of the depth prediction based on the overall detection area and the unit detection area includes: The ratio of the overall detection area to the unit detection area is calculated, and the ratio is used as the regression period.
8. A device for predicting the maximum corrosion depth of a tank bottom plate, characterized in that: include: An acquisition module, used for acquiring corrosion depth data of a tank bottom plate, an overall detection area of the tank bottom plate, and a unit detection area of a single plate of the tank bottom plate; A determination module, configured to calculate a cumulative probability corresponding to the corrosion depth data, and determine a target generalized extreme value distribution of the corrosion depth data based on the cumulative probability; A calculation module, used for calculating the target cumulative probability corresponding to the maximum corrosion depth of the tank bottom plate based on the overall detection area and the unit detection area; The prediction module is used to determine the target corrosion depth corresponding to the target cumulative probability based on the target generalized extreme value distribution, and use the target corrosion depth as the maximum corrosion depth of the tank bottom plate.
9. An electronic device comprising a memory, a processor and a computer program, wherein the computer program is stored in the memory and can be run on the processor, wherein: When the processor executes the program, the method for predicting the maximum corrosion depth of the tank bottom plate according to any one of claims 1 to 7 is implemented.
10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for predicting the maximum corrosion depth of the tank bottom plate according to any one of claims 1 to 7 is implemented.
Citation Information
Cited By
Prediction method for maximum corrosion depth of metal matrix at damaged part of coating of earthing storage tank
CN122433356A
