Method for determining the chloride ion permeability of concrete in an underground structure

Abstract

The application discloses a method for determining the chloride ion permeability of concrete in an underground structure, comprising the following steps: S1, collecting recycled aggregate concrete component data and permeability indexes; S2, dividing the data collected in step S1 into a training set and a test set; S3, establishing a chloride ion permeability prediction model by using an adaptive neural fuzzy inference system; S4, inputting the training set in step S2 into the chloride ion permeability prediction model in step S3, and optimizing and adjusting the chloride ion permeability prediction model parameters by using a chaos-based glowworm algorithm; S5, inputting the test set in step S2 into the optimized chloride ion permeability prediction model, testing the chloride ion permeability prediction error, and obtaining a trained chloride ion permeability prediction model; and S6, predicting the chloride ion permeability by using the trained chloride ion permeability prediction model. The method overcomes the defects of long time period and high test cost of the prior art, and realizes accurate prediction of the chloride ion permeability of concrete.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of civil engineering material performance test methods, and more particularly, to a method for determining the chloride ion permeability of concrete in underground structures. BACKGROUND

[0002] In the process of modern city construction, the emergence of underground structures solves the problem of insufficient urban development space. Especially for some developed coastal cities, the emergence of underground structures provides a new direction for the further development of the city. Underground structures are now widely used in urban regional construction, but the lack of aggregate sources during construction has become a problem that needs to be solved. The emergence of recycled aggregate concrete solves this problem. Recycled aggregate concrete saves natural resources and reduces environmental pollution by recycling construction waste, achieving sustainable development of building technology, and has good development prospects. At present, durability is a key factor for the popularization and application of recycled aggregate concrete, and chloride ion permeability is an important indicator for evaluating the durability of recycled aggregate concrete. Chloride ion permeability reflects the compactness of concrete and the ability to resist the intrusion of external media into the interior of the concrete. The lower the chloride ion permeability, the higher the compactness of the concrete, the stronger the ability to resist chloride ion erosion, and the stronger the durability of the steel structure inside the concrete. Chloride ion permeability is closely related to the composition, proportion, water-cement ratio, curing period and other influencing factors of concrete. The current methods for determining chloride ion permeability are mainly experimental methods and numerical methods, which involve setting up recycled aggregate concrete samples with different proportions and components for testing, and selecting the sample with the lowest chloride ion permeability coefficient as the final proportion. These methods have a long time period and high testing costs. Therefore, a method for determining the chloride ion permeability of concrete in underground structures is needed.

[0003] A loading device and testing method for testing the chloride ion permeability of concrete are disclosed in Chinese Patent No. CN101718671A, published on June 2, 2010. The patent states: "A loading device for testing the chloride ion permeability of concrete is disclosed, which includes a longitudinal loading system and a lateral loading system. The testing method for testing the chloride ion permeability using the loading device is also disclosed. The present application solves the problem that the conventional testing device for testing the chloride ion permeability of concrete cannot load the tested sample, cannot simulate the actual load state of the in-service concrete, and cannot determine the chloride ion permeability of concrete under pressure load." The new device and testing method proposed in this patent are based on actual simulation tests and calculations, which can provide a reference for construction to some extent. However, this calculation method requires pouring a model of the concrete with the specified proportion and components first, and then calculating the chloride ion permeability, which requires setting up multiple groups of samples with different proportions and components, resulting in a long time period and high testing costs. SUMMARY

[0004] The application provides a method for determining the chloride ion permeability of concrete in an underground structure, which overcomes the defects of long time period and high test cost of the prior art and realizes accurate prediction of the chloride ion permeability of concrete.

[0005] To solve the above technical problems, the technical scheme of the application is as follows:

[0006] A method for determining the chloride ion permeability of concrete in an underground structure, comprising the following steps:

[0007] S1: collecting component data and permeability indexes of recycled aggregate concrete;

[0008] S2: dividing the data collected in step S1 into a training set and a test set;

[0009] S3: establishing a chloride ion permeability prediction model by using an adaptive neuro-fuzzy inference system;

[0010] S4: inputting the training set in step S2 into the chloride ion permeability prediction model in step S3, and optimizing and adjusting the parameters of the chloride ion permeability prediction model by using a glowworm swarm algorithm based on chaos;

[0011] S5: inputting the test set in step S2 into the optimized chloride ion permeability prediction model, testing the prediction error of the chloride ion permeability, and obtaining a trained chloride ion permeability prediction model;

[0012] S6: predicting the chloride ion permeability by using the trained chloride ion permeability prediction model.

[0013] Preferably, the component data of recycled aggregate concrete in step S1 comprises the water content w, cement content c, coarse recycled aggregate RA, sand content S, pozzolanic material content Pm, water-cement ratio w / c, curing age T, particle density D and water absorption rate Wa of recycled aggregate concrete.

[0014] Preferably, the permeability index in step S1 is the charge amount entering the recycled aggregate concrete per unit time.

[0015] Preferably, the training set in step S2 is part of the component data and permeability indexes of recycled aggregate concrete collected in step S1, wherein the component data of recycled aggregate concrete is used as the input data of the chloride ion permeability prediction model, and the permeability index is used as the output data of the chloride ion permeability prediction model.

[0016] Preferably, in step S2, the remaining component data and permeability indexes of recycled aggregate concrete except the training set form a test set, and the data types of the training set and the test set are the same.

[0017] Preferably, the adaptive neuro-fuzzy inference system is used to establish the chloride ion permeability prediction model in step S3, which comprises five layers, each of which is defined as follows:

[0018] (1) The first layer defines the fuzzy membership function of each node, and the calculation formula is as follows:

[0019]

[0020] wherein, I m represents the mth input data, I m k represents the kth rule corresponding to the mth input data, represents the membership function of the kth rule corresponding to the mth input data, O 1，k represents the output data of the kth rule corresponding to the first layer of the chloride ion permeability prediction model;

[0021] (2) The second layer calculates the trigger strength w k of the kth rule, and the calculation formula is as follows:

[0022]

[0023] wherein, O 2，k represents the output data of the kth rule corresponding to the second layer of the chloride ion permeability prediction model;

[0024] (3) The third layer calculates the average trigger strength w of the kth rule, and the calculation formula is as follows:

[0025]

[0026] wherein, O 3，k represents the output data of the kth rule corresponding to the third layer of the chloride ion permeability prediction model;

[0027] (4) The fourth layer calculates the influence of the kth rule on the model output, and the calculation formula is as follows:

[0028]

[0029]

[0030] wherein, O 4，k represents the output data of the kth rule corresponding to the fourth layer of the chloride ion permeability prediction model, f k is the kth node function, p m k is the coefficient of the kth rule corresponding to the mth input data, r k is the bias coefficient of the kth node function;

[0031] (5) the fifth layer calculates the output data O5(P) of the chloride ion permeability prediction model, i.e. the permeability index prediction value, as follows:

[0032]

[0033] In the formula, O5(P) represents the output data of the chloride ion permeability prediction model.

[0034] Preferably, the step S4 optimizes and adjusts the parameters of the chloride ion permeability prediction model by using the chaos-based glowworm swarm optimization algorithm, and the parameters of the chloride ion permeability prediction model are specifically the set of p m k and r k .

[0035] Preferably, the step S4 uses the chaos-based glowworm swarm optimization algorithm, specifically integrates the chaos mapping and Lévy meta-heuristic algorithm mechanism into the typical glowworm algorithm, wherein the chaos mapping is to generate a chaotic number between 0 and 1 by using a chaotic number generator, and to perform the initialization, selection, crossover and mutation operations on the population by using the chaotic number; and the Lévy meta-heuristic algorithm is used to accelerate the iteration process.

[0036] Preferably, the step S4 optimizes and adjusts the parameters of the chloride ion permeability prediction model, and specifically includes:

[0037] (a) determining the number of initial populations, performing population initialization, and setting the maximum number of iterations;

[0038] (b) calculating the attraction parameter of the tthiteration by using the Gaussian chaos mapping, and the calculation formula is as follows:

[0039]

[0040] In the formula, t represents the iteration number, t represents the chaotic number of the tthiteration, γ is the absorption coefficient, β0 is a random number generated by the normal distribution of [0, 1], and β t is the attraction parameter of the tthiteration, r ij is the distance between the vector P i and P j .

[0041] (c) calculating the Lévy distribution L(P t n of the tthiteration population P t n by using the Lévy meta-heuristic algorithm, and the calculation formula is as follows:

[0042]

[0043]

[0044]

[0045]

[0046] wherein τ is a constant;

[0047] (d) calculating the candidate parameter P of the t = t + 1th iteration n t+1 , n = 1,..., 100, and the calculation formula is as follows:

[0048]

[0049] ε = rand - 1 / 2

[0050] α t = α0θ t

[0051] wherein and represent the i-th and j-th estimated values of the tth iteration parameter P, α t represents the weighting coefficient at the tth iteration, ε represents a random vector number estimated by a uniform distribution and a Gaussian distribution, rand represents a random number generated by a regular distribution of [0, 1], and θ represents a random reduction coefficient, 0 < θ < 1;

[0052] (e) calculating the root mean square error Error (P n t+1 ) as follows:

[0053]

[0054] wherein O'5 is an actual chloride ion permeability value, and the candidate parameter P n t+1 is selected when Error (P n t+1 is the minimum;

[0055] (f) setting a stop condition, and when the stop condition is reached or the maximum number of iterations is reached, outputting the optimal model parameter, ending the algorithm, and saving the model.

[0056] Preferably, τ is 1.5.

[0057] Compared with the prior art, the technical scheme of the present application has the beneficial effects that:

[0058] The present application aims at the defects of high test cost and long time period in the prior art, and provides a method for determining the chloride ion permeability of concrete in an underground structure, which establishes a model by using an adaptive fuzzy neural inference system and a glowworm algorithm based on chaos theory, optimizes parameters to improve the calculation efficiency and the calculation accuracy, and establishes the relationship between the concrete permeability and the concrete components, so that the chloride ion permeability of recycled aggregate concrete under various influencing factors can be quickly determined, and the whole method can realize the accurate prediction of the chloride ion permeability of recycled aggregate concrete without assuming a model theory. BRIEF DESCRIPTION OF DRAWINGS

[0059] Figure 1 It is a method flowchart of the present application.

[0060] Figure 2 It is a comparison diagram of the prediction results and the measured results in the embodiment by using the method of the present application. DETAILED DESCRIPTION

[0061] The drawings are only used for illustrative description, and cannot be understood as a limitation on the present patent;

[0062] In order to better illustrate the present embodiment, some components in the drawings may be omitted, enlarged or reduced, and do not represent the size of the actual product;

[0063] For those skilled in the art, it is understandable that some well-known structures and their descriptions in the drawings may be omitted.

[0064] The technical solutions of the present application will be further described below in combination with the drawings and embodiments.

[0065] Embodiment 1

[0066] The present embodiment provides a method for determining the chloride ion permeability of concrete in an underground structure, as shown in the figure, comprising the following steps: Figure 1 S1: collecting recycled aggregate concrete component data and permeability indicators;

[0067] S1: collecting recycled aggregate concrete component data and permeability indicators;

[0068] S2: dividing the data collected in step S1 into a training set and a test set;

[0069] S3: establishing a chloride ion permeability prediction model by using an adaptive neural fuzzy inference system;

[0070] S4: inputting the training set in step S2 into the chloride ion permeability prediction model in step S3, and optimizing and adjusting the chloride ion permeability prediction model parameters by using a glowworm algorithm based on chaos;

[0071] S5: inputting the test set in step S2 into the optimized chloride ion permeability prediction model to test the chloride ion permeability prediction error, and obtaining the trained chloride ion permeability prediction model;

[0072] S6: predicting the chloride ion permeability by using the trained chloride ion permeability prediction model.

[0073] In the embodiment, the data collected comes from 83 groups of experimental results involved in multiple literatures, and the component data of the recycled aggregate concrete in step S1 includes the water content w, the cement content c, the coarse recycled aggregate RA, the sand content S, the pozzolanic material content Pm, the water-cement ratio w / c, the curing age T, the particle density D and the water absorption rate Wa of the recycled aggregate concrete.

[0074] In the embodiment, the water content w refers to the ratio of the weight of water contained in the recycled aggregate concrete to the total weight of the recycled aggregate concrete, the minimum value in the collected data is 157.5, the maximum value is 226.83, and the average value is 205.42 (unit: kg / m 3 );the cement content c refers to the ratio of the weight of cement in the recycled aggregate concrete to the total weight of the recycled aggregate concrete, the minimum value in the collected data is 175.5, the maximum value is 553.5, and the average value is 371.57 (unit: kg / m 3 );the coarse recycled aggregate RA refers to the aggregate prepared from waste concrete, the minimum value in the collected data is 832.14, the maximum value is 1080, and the average value is 1011.46 (unit: kg / m 3 );the sand content S refers to the ratio of the weight of sand contained in the recycled aggregate concrete to the total weight of the recycled aggregate concrete, the minimum value in the collected data is 479.0, the maximum value is 787, and the average value is 665.86 (unit: kg / m 3 );the pozzolanic material content Pm refers to the ratio of the weight of pozzolanic material contained in the recycled aggregate concrete to the total weight of the recycled aggregate concrete, the minimum value in the collected data is 11.13, the maximum value is 227.5, and the average value is 101.09 (unit: kg / m 3); the water-cement ratio w / c refers to the weight ratio of water to cement in the recycled aggregate concrete, the minimum value in the collected data is 0.35, the maximum value is 0.55, and the average value is 0.47; the curing age T refers to the time experienced by the recycled aggregate concrete after mixing and forming after a certain curing, the minimum value in the collected data is 1.0, the maximum value is 365, and the average value is 60.97 (unit: days); the particle density D refers to the density of various particles in the recycled aggregate concrete, the minimum value in the collected data is 2.34, the maximum value is 4.8, and the average value is 2.59; and the water absorption rate Wa refers to the ratio of the weight of water absorbed by the standard cubic recycled aggregate concrete to the total weight of the recycled concrete, the minimum value in the collected data is 3.89%, the maximum value is 7.06%, and the average value is 5.04%

[0075] The permeability index in the step S1 refers to the amount of electric charges entering the recycled aggregate concrete per unit time, representing the size of the chloride ion permeability of the recycled aggregate concrete.

[0076] The training set in the step S2 is 66 groups of data selected from the component data and the permeability index of the recycled aggregate concrete collected in the step S1, wherein the component data of the recycled aggregate concrete is taken as the input data of the chloride ion permeability prediction model, and the permeability index is taken as the output data of the chloride ion permeability prediction model.

[0077] In addition to the training set, the remaining 17 groups of component data and permeability index of the recycled aggregate concrete in the step S2 form a test set, which has the same data type as the training set and is used to verify the prediction accuracy of the chloride ion permeability prediction model.

[0078] The chloride ion permeability prediction model is established by using the adaptive neuro-fuzzy inference system in the step S3, which specifically includes five layers, and each layer is defined as follows:

[0079] (1) The first layer defines the fuzzy membership function of each node, and in this embodiment, the input data is sequentially defined as the 1st to 9th nodes according to the enumeration order of the input data in the step two. Two rules are set for each node to generate the fuzzy membership function, and the calculation formula is as follows:

[0080]

[0081] In the formula, I m represents the mth input data, I m k represents the kth rule corresponding to the mth input data, represents the membership function of the kth rule corresponding to the mth input data, O 1，k represents the output data of the kth rule corresponding to the first layer of the chloride ion permeability prediction model, and in this embodiment, the Gaussian membership function is adopted: where c=0, σ=1. O 1，1 , O 1，2 represents the output data of the first layer corresponding to the first and second rules;

[0082] (2) The second layer calculates the trigger strength w k of the kth rule. In this embodiment, the nine fuzzy membership functions corresponding to each rule of the first layer are multiplied to obtain the trigger strength w k of the rule. The calculation formula is as follows:

[0083]

[0084] where O 2，k represents the output data of the kth rule corresponding to the second layer of the chloride ion permeability prediction model;

[0085] (3) The third layer calculates the average trigger strength w of the kth rule. The calculation formula is as follows:

[0086]

[0087] where O 3，k represents the output data of the kth rule corresponding to the third layer of the chloride ion permeability prediction model;

[0088] (4) The fourth layer calculates the influence of the kth rule on the model output, and the calculation formula is as follows:

[0089]

[0090]

[0091] where O 4，k represents the output data of the kth rule corresponding to the fourth layer of the chloride ion permeability prediction model, f k is the kth node function, p m k is the coefficient of the kth rule corresponding to the mth input data, r k is the bias coefficient of the kth node function;

[0092] (5) The fifth layer calculates the output data O5(P) of the chloride ion permeability prediction model, i.e., the predicted value of the permeability index, as follows:

[0093]

[0094] where O5(P) represents the output data of the chloride ion permeability prediction model.

[0095] The step S4 utilizes the chaos-based glowworm algorithm to optimize and adjust the chloride ion permeability prediction model parameters, and the chloride ion permeability prediction model parameters are specifically p m k and r k , and the parameters of the chloride ion permeability prediction model are denoted as a vector

[0096] The step S4 utilizes the chaos-based glowworm algorithm, specifically a chaos mapping and a Lévy meta-heuristic algorithm mechanism are integrated into a typical glowworm algorithm, wherein the chaos mapping is to generate a chaos number between 0 and 1 by using a chaos number generator, and the chaos number is used to perform initialization, selection, crossover and mutation operations on the population; and the Lévy meta-heuristic algorithm is used to accelerate the iteration process.

[0097] The step S4 optimizes and adjusts the chloride ion permeability prediction model parameters, and specifically includes:

[0098] (a) t = 0, 100 initial candidate populations are randomly set, that is, 100 groups of candidate parameters P, denoted as P n (n = 1,..., 100), population initialization is performed, and the maximum number of iterations is set to 1000;

[0099] (b) the attraction parameter of the tthiteration is calculated by using the Gaussian chaos mapping, and the calculation formula is as follows:

[0100]

[0101] In the formula, t represents the iteration number, represents the tthchaos number, γ is an absorption coefficient, and the value is 1.2; β0is a random number generated by a normal distribution of [0, 1], and in the embodiment, β0takes a value of 0.2; β t is the attraction parameter of the tthiteration, and r ij is the distance between the vector P i and P j

[0102] (c) the Lévy distribution L(P t n ) of the tthiteration population P t n is calculated by using the Lévy meta-heuristic algorithm, and the calculation formula is as follows:

[0103]

[0104]

[0105] ​

[0106]

[0107] where τ is a constant;

[0108] (d) Calculate the candidate parameters P of the t = t + 1 iteration n t+1 , n = 1,..., 100, the calculation formula is as follows:

[0109]

[0110] ε = rand - 1 / 2

[0111] α t = α0θ t

[0112] wherein, and represent the i-th and j-th estimated values of the t-th iteration parameter P, α t represents the weighting coefficient at the t-th iteration, ε represents the random vector number estimated by the uniform distribution and the Gaussian distribution, rand represents the random number generated by the regular distribution of [0, 1], θ represents the random reduction coefficient, 0 < θ < 1;

[0113] (e) Calculate the root mean square error Error (P n t+1 ), as follows:

[0114]

[0115] wherein O'5 is the actual chloride ion permeability value, the candidate parameter P n t+1 is selected, which minimizes the error Error (P n t+1 is the optimal model parameter P of the t + 1 iteration;

[0116] (f) If the optimal model output value and the actual value error is within 10% or reaches the maximum iteration number 1000, the iteration can be stopped, the current optimal model parameter P = {0.56, 0.43, 0.1, 0.88, 0.96, 0.07, 0.34, 0.58, 0.68, 0.18, 0.22, 0.58, 0.52, 0.42, 0.68, 0.97, 0.56, 0.05, 0.12, 0.48, 0.93} is output, and the model is saved. Otherwise, return to b) re-iteration until convergence.

[0117] The τ takes 1.5.

[0118] In the embodiment, the test set data is input into the chloride ion permeability prediction model, the chloride ion permeability is predicted, and the predicted value and the measured value are plotted in the same coordinate system to obtain a comparison diagram of the predicted value and the measured value, as shown in Figure 2 The prediction error rate of the model is only 12%.

[0119] The same or similar reference signs correspond to the same or similar components;

[0120] The terms describing the positional relationship in the drawings are only used for illustrative description, and should not be understood as a limitation on the patent;

[0121] Obviously, the above embodiments of the present application are only examples for clearly illustrating the present application, and are not intended to limit the implementation modes of the present application. Based on the above description, other different forms of changes or variations can be made by those skilled in the art. Here, it is not necessary and also impossible to enumerate all the implementation modes. Any modification, equivalent replacement and improvement made within the spirit and principle of the present application should be included in the protection scope of the claims of the present application.

Claims

1. A method for determining the chloride ion permeability of concrete in an underground structure, characterized by, The method comprises the following steps: S1: collecting recycled aggregate concrete component data and permeability index; S2: dividing the data collected in step S1 into a training set and a test set; S3: establishing a chloride ion permeability prediction model by using an adaptive neuro-fuzzy inference system; S4: inputting the training set in step S2 into the chloride ion permeability prediction model in step S3, and optimizing and adjusting the parameters of the chloride ion permeability prediction model by using a chaos-based glowworm swarm optimization algorithm; S5: inputting the test set in step S2 into the optimized chloride ion permeability prediction model, testing the chloride ion permeability prediction error, and obtaining a trained chloride ion permeability prediction model; S6: predicting the chloride ion permeability by using the trained chloride ion permeability prediction model. In step S4, the parameters of the chloride ion permeability prediction model are optimized and adjusted, and the method comprises the following steps: (a) determining the number of initial populations, performing population initialization, and setting the maximum number of iterations; (b) calculating the attraction force parameters of the t-th iteration by using Gaussian chaos mapping, and the calculation formula is as follows: In the formula, t represents the number of iterations. Let γ represent the chaos number at time t, γ be the absorption coefficient, β0 be a random number generated by a normal distribution in the range [0,1], and β be the number of random numbers generated by a normal distribution in the range [0,1]. t Let r be the attraction parameter for the t-th iteration. ij For vector P i With P j Distance between (c) Calculate the population P of the tth iteration using the Lévy metaheuristic algorithm t n of the Lévy distribution L(P t n ), whose formula is as follows: wherein τ is a constant; (d) calculate the candidate parameters P for the t = t + 1 iteration n t+1 , n = 1,..., 100, with the following formula: ε=rand-1 / 2 α t = a0θ t In the formula, and represents the i-th and j-th estimated values of the t-th iteration parameter P, and α t represents a weighting coefficient at the t-th iteration, ε represents a random vector number estimated by a uniform distribution and a Gaussian distribution, rand represents a random number generated by a regular distribution of [0, 1], and θ represents a random reduction coefficient, 0 < θ < 1. (e) calculating the root mean square error Error(P n t+1 ), as follows: where O'5 is the actual chloride ion permeability value, and Error(P n t+1 ) the candidate parameter P n t+1 is the optimal model parameter P for the t+1 iteration; (f) setting a stop condition, outputting the optimal model parameters when the stop condition or the maximum number of iterations is reached, ending the algorithm, and saving the model.

2. The method of determining the chloride ion permeability of concrete in a subterranean structure according to claim 1, characterized in that, In step S1, the recycled aggregate concrete component data comprises the water content w, the cement content c, the coarse recycled aggregate RA, the sand content S, the pozzolanic material content Pm, the water-cement ratio w / c, the curing age T, the particle density D, and the water absorption rate Wa of the recycled aggregate concrete.

3. The method of determining the chloride ion permeability of concrete in a subterranean structure according to claim 1, wherein In step S1, the permeability index is the amount of electric charge entering the recycled aggregate concrete per unit time.

4. The method of determining the chloride ion permeability of concrete in a subterranean structure according to claim 1, wherein In step S2, the training set is part of the recycled aggregate concrete component data and the permeability index collected in step S1, wherein the recycled aggregate concrete component data is used as the input data of the chloride ion permeability prediction model, and the permeability index is used as the output data of the chloride ion permeability prediction model.

5. The method of determining the chloride ion permeability of concrete in a subterranean structure according to claim 4, wherein In step S2, the remaining recycled aggregate concrete component data and permeability index except the training set form the test set, and the data type of the test set is the same as that of the training set.

6. The method of determining the chloride ion permeability of concrete in a subterranean structure according to claim 1, wherein In step S3, the chloride ion permeability prediction model is established by using the adaptive neuro-fuzzy inference system, and the model comprises five layers, and each layer is defined as follows: (1) the first layer defines the fuzzy membership function of each node, and the calculation formula is as follows: In the formula, I m represents the mth input data, I m k represents the kth rule corresponding to the mth input data, represents the membership function of the kth rule corresponding to the mth input data, O 1,k represents the output data of the kth rule corresponding to the first layer of the chloride ion permeability prediction model; (2) The second layer calculates the trigger strength w of the kth rule k The calculation formula is as follows: In the formula, O 2,k represents the output data of the kth rule corresponding to the second layer of the chloride ion permeability prediction model (3) The third layer calculates the average trigger strength of the kth rule The calculation formula is as follows: In the formula, O 3,k represents the output data of the kth rule corresponding to the third layer of the chloride ion permeability prediction model (4) the fourth layer calculates the influence of the kth rule on the model output, and the calculation formula is as follows: In the formula, O 4,k Output data of the kth rule corresponding to the fourth layer of the chloride ion permeability prediction model, f k The kth node function, p m k The coefficient of the kth rule corresponding to the mth input data, r k The bias coefficient of the kth node function; (5) the fifth layer calculates the output data O5(P) of the chloride ion permeability prediction model, i.e. the predicted value of the permeability index, as follows: wherein O5(P) represents the output data of the chloride ion permeability prediction model.

7. The method of determining the chloride ion permeability of concrete in a subterranean structure according to claim 6, wherein The step S4 utilizes the chaos-based glowworm swarm optimization algorithm to optimize and adjust the chloride ion permeability prediction model parameters, and the chloride ion permeability prediction model parameters are specifically a set of p m k and r k .

8. The method of determining the chloride ion permeability of concrete in a subterranean structure according to claim 7, wherein In step S4, the chaos-based glowworm swarm optimization algorithm is used, and the algorithm is obtained by integrating the chaos mapping and the Lévy meta-heuristic algorithm mechanism into the typical glowworm swarm optimization algorithm, wherein the chaos mapping is used to generate chaos numbers between 0 and 1 by using a chaos number generator, and the chaos numbers are used to perform the initialization, selection, crossover, and mutation operations on the population; and the Lévy meta-heuristic algorithm is used to accelerate the iteration process.

9. The method of determining the chloride ion permeability of concrete in a subterranean structure according to claim 8, wherein The value of τ is 1.5.

Images