A method, apparatus, and storage medium for predicting the frost resistance of concrete containment structures with pipe openings based on the PSO-BP algorithm.

Abstract

This invention discloses a method, apparatus, and storage medium for predicting the frost resistance of containment concrete with pipe openings based on the PSO-BP algorithm. The method includes: preprocessing sample data by using the mass parameter of spalling material from a single-sided freeze-thaw cycle of the containment concrete with pipe openings as an index of its frost resistance; constructing a prediction model based on the PSO-BP algorithm; setting the number of particle swarm iterations and the population size; obtaining individual optimal values ​​and a global optimal value through calculation and iterative updates; assigning the global optimal value to a BP neural network; training the assigned BP neural network based on the obtained measurement dataset to obtain the final prediction model, which predicts the mass loss of any pipe diameter under any number of freeze-thaw cycles. Compared to traditional surface fitting methods, this invention provides more accurate and stable prediction results and can serve as an effective tool for rapidly predicting the frost resistance of containment concrete with pipe openings.

Description

Technical Field

[0001] This invention belongs to the field of building material durability prediction technology, and relates to a method, device and storage medium for predicting the frost resistance of concrete containment structures with pipe openings based on the PSO-BP algorithm. Background Technology

[0002] Nuclear power accounts for a large portion of the world's electricity. With the continuous growth of electricity demand, since the 20th century, some countries have begun to use nuclear power as a supplement to conventional power. The containment concrete of nuclear power plants is subject to corrosion from environmental factors such as chloride ions and sulfates, and coastal containment concrete is also affected by freeze-thaw cycles. The damage caused by freeze-thaw cycles not only affects the aesthetics of the concrete structure but also reduces its mechanical properties and durability. Generally, the containment concrete structure consists of a dome, ring beams, prestressed walls, and reinforced concrete foundations. Due to usage and the installation of equipment pipelines, pipes are often embedded in the containment concrete structure for pre-reserved openings. Obviously, the area around the pipes and openings is a weak point in the containment concrete, which reduces its mechanical properties and durability. Therefore, predicting the freeze-thaw resistance of containment concrete with pipe openings (i.e., containment concrete with pre-reserved pipe openings) becomes very important.

[0003] Predicting the frost resistance of concrete containment structures with pipe openings typically involves extensive experimental data and surface fitting methods. This approach not only results in a large number of experiments and a massive computational burden for data processing and fitting, but also leads to a significant average absolute percentage error between the predicted and experimental results. Furthermore, the differences between actual freeze-thaw cycles and experimental conditions amplify the errors in the predictions. Therefore, a smart method with lower data dependence and stable prediction performance is needed to predict the frost resistance of concrete containment structures with pipe openings, reflecting their true frost resistance and providing valuable guidance for related engineering designs. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method, device and storage medium for predicting the frost resistance of containment concrete with pipe openings based on the PSO-BP algorithm, thereby solving the problems of large computational load and large average absolute percentage error caused by surface fitting.

[0005] To achieve the above objectives, the present invention relates to a method, apparatus, and storage medium for predicting the frost resistance of concrete containment structures with pipe openings based on the PSO-BP algorithm, comprising the following steps:

[0006] Step 1: The mass parameters of the spalling material after single-sided freeze-thaw of the containment concrete with pipe openings are used as the freeze-thaw resistance index of the containment concrete with pipe openings. Based on the selected freeze-thaw resistance index of the containment concrete, sample data belonging to the index are collected on multiple containment concrete specimens with pipe openings of different diameters.

[0007] Step 2: Perform sample data preprocessing to obtain preprocessed measurement data;

[0008] Step 3: Construct a prediction model for the frost resistance of concrete containment structures with pipe openings based on the PSO-BP algorithm;

[0009] Step 4: Set the number of particle swarm iterations and the population size. Through calculation and iterative updates, obtain the individual optimal value and the global optimal value. Assign the global optimal value to the BP neural network.

[0010] Step 5: Based on the dataset of measured values ​​obtained in Step 2, take the number of freeze-thaw cycles and the diameter of the pipe with the opening as two influencing factors as input variables, and take the measured value of the total mass of the actual unit surface area of ​​the specimen as the output variable to train the BP neural network after the assignment in Step 4.

[0011] Step 6: Test the preprocessed sample dataset with the trained prediction model for the frost resistance of containment concrete with pipe openings based on the PSO-BP algorithm, and determine the R-value between the prediction model's prediction results and the measured dataset. 2 Whether it is greater than 95%; if the R-value of the prediction model prediction result is greater than or equal to the measured value dataset. 2 If the value is less than 95%, adjust the number of particle swarm optimization iterations to obtain new optimal individual values ​​and new optimal population values, until the R-value between the prediction model's prediction and the measured dataset is found to be within acceptable limits. 2 If the accuracy is greater than 95%, the final prediction model is obtained. Using the final prediction model, the mass loss of any pipe diameter under any number of freeze-thaw cycles can be predicted.

[0012] Furthermore, in step 1, using the mass of spalling material from a single-sided freeze-thaw cycle of the concrete with pipe openings as an index of the concrete's freeze-thaw resistance means that the mass of spalling material from a single-sided freeze-thaw cycle of the concrete specimen with pipe openings is converted into an index of the concrete's freeze-thaw resistance. The collected data is the mass of spalling material on the surface of the specimen after freeze-thaw cycles.

[0013] Furthermore, the conversion method is as follows:

[0014] μ s =μ f -μ b (1)

[0015] A = b 2 -πd 2 / 4 (2)

[0016]

[0017] In the formula, μ s The mass (g) of the flaking material on the specimen surface is accurate to 0.01g; μ b The mass of the filter paper (g) is accurate to 0.01g; μ f The total mass (g) of the dried filter paper and the residue from the test specimen, accurate to 0.01 g; ∑μ s Let M be the total mass of the surface spalling material of the specimen after N cumulative single-sided freeze-thaw cycles, and M be the total mass of spalling material per unit test surface area of ​​a single specimen after N freeze-thaw cycles (g / m²). 2 A represents the surface area (mm²) of a single specimen's test surface. 2 b is the side length (mm) of the test surface of the specimen; d is the diameter (mm) of the steel pipe in the concrete containment structure with pipe opening.

[0018] Furthermore, the pretreatment in step 2 includes: taking the arithmetic mean of the total mass of spalling material per unit test surface area of ​​multiple concrete specimens with pipe openings of the same specification in the same group as the measured value of the total mass of spalling material per unit surface area of ​​the specimens in that group.

[0019] Furthermore, in step 3, the construction of the prediction model for the frost resistance of containment concrete with pipe openings based on the PSO-BP algorithm specifically includes the following steps:

[0020] Step 3-1: Determine the topology of the BP network;

[0021] Step 3-2: Generate the initial population, and initialize particle velocity, position, and individual and global optimum.

[0022] Step 3-3: Calculate the error and use the calculated error as the particle fitness to update the individual best and the global best.

[0023] Steps 3-4: Update the example speed and location;

[0024] Step 3-5: Repeat steps 3-3 to 3-4 until the termination condition is met;

[0025] Steps 3-6: Assign initial weights and thresholds to the BP network based on the particle swarm positions;

[0026] Steps 3-7: Calculate the error, and update the weights and thresholds;

[0027] Step 3-8, repeat step 3-7 until the termination condition is met;

[0028] Steps 3-9: Output the predicted values.

[0029] Furthermore, in step 3-3, the particle updates its velocity and position through individual optimality and global optimality, using the following update formula:

[0030]

[0031]

[0032] In the formula, V i P represents particle velocity; ω represents relation weight; d = 1, 2, ..., D, where D is the dimension of the search space; k is the current iteration number; P i For the individual best; P g For global best; X i denoted as the particle position; c1 and c2 are non-negative constants; r1 and r2 are random numbers distributed between [0,1].

[0033] Furthermore, in steps 3-5, the number of iterations for the termination condition reaches its maximum value.

[0034] Furthermore, in steps 3-8, the termination condition is that the network prediction error is less than the given accuracy, or the number of iterations reaches the maximum value.

[0035] As another aspect of the present invention, an apparatus for predicting the frost resistance of concrete with pipe openings in containment structures is also provided. The apparatus includes a data acquisition module for using the mass parameters of the spalling material after single-sided freeze-thaw of the concrete with pipe openings in containment structures as an index of the frost resistance of the concrete with pipe openings in containment structures. Based on the selected frost resistance index of the concrete with pipe openings in containment structures, the apparatus collects the index and obtains sample data for each group of multiple concrete specimens with pipe openings in containment structures of the same pipe diameter.

[0036] The data preprocessing module is used to preprocess sample data to obtain a preprocessed dataset of measured values.

[0037] The prediction model building module is used to build a prediction model for the frost resistance of containment concrete with pipe openings based on the PSO-BP algorithm.

[0038] The prediction model training module uses the set number of particle swarm iterations and population size to obtain individual optimal values ​​and global optimal values ​​through calculation and iterative updates. The global optimal value is then assigned to the BP neural network. Based on the obtained measurement dataset, the number of freeze-thaw cycles and the diameter of the pipe with the opening are used as two influencing factors as input variables, and the total mass of the actual unit surface area of ​​the specimen is used as the output variable to train the BP neural network after the assignment in step 4.

[0039] The mass loss prediction module uses a trained prediction model based on the PSO-BP algorithm for the frost resistance of containment concrete with pipe openings to test the preprocessed sample dataset, and determines the R-value between the prediction model's prediction results and the measured dataset. 2 Whether it is greater than 95%; if the R-value of the prediction model prediction result is greater than or equal to the measured value dataset. 2 If the value is less than 95%, adjust the number of particle swarm optimization iterations to obtain new optimal individual values ​​and new optimal population values, until the R-value between the prediction model's prediction and the measured dataset is found to be within acceptable limits. 2 If the accuracy is greater than 95%, the final prediction model is obtained. Using the final prediction model, the mass loss of any pipe diameter under any number of freeze-thaw cycles can be predicted.

[0040] As another aspect of the present invention, a computer storage medium storing a computer program is also provided, characterized in that, when the computer program is executed by a processor, it implements the above-described method for predicting the frost resistance of containment concrete with pipe openings based on the PSO-BP algorithm.

[0041] In summary, compared with the prior art, the above-described technical solutions conceived by this invention can achieve the following beneficial effects:

[0042] (1) The method of predicting the frost resistance of concrete with pipe opening based on PSO-BP algorithm of the present invention adopts an intelligent method with less data dependence and stable prediction effect. Compared with the traditional fitting surface method, it can obtain more accurate and stable prediction results and can be used as an effective tool for quickly predicting the frost resistance of concrete with pipe opening. Attached Figure Description

[0043] Figure 1 This is a flowchart of a preferred embodiment of the present invention;

[0044] Figure 2 This is a schematic diagram of a preferred concrete specimen of the present invention;

[0045] Figure 3 This is a graph showing the temperature change over time during a single freeze-thaw cycle according to a preferred embodiment of the present invention.

[0046] Figure 4 This is the topology of a BP network according to a preferred embodiment of the present invention;

[0047] Figure 5 This is a flowchart of the PSO-BP algorithm of a preferred embodiment of the present invention;

[0048] Figure 6 This is a comparison chart of the test results and predicted results of M under the same pipe diameter and different freeze-thaw cycle numbers N according to a preferred embodiment of the present invention.

[0049] Figure 7 This is a structural diagram of a device for predicting the frost resistance of a containment concrete structure with a pipe opening, according to a preferred embodiment of the present invention. Detailed Implementation

[0050] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described examples are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0051] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0052] Example 1:

[0053] Please refer to Figures 1-5 Example 1 provides a method for predicting the frost resistance of concrete containment structures with pipe openings based on the PSO-BP algorithm, comprising the following steps:

[0054] Step 1: The mass parameters of the spalling material after single-sided freeze-thaw of the containment concrete with pipe openings are used as the freeze-thaw resistance index of the containment concrete with pipe openings. Based on the selected freeze-thaw resistance index of the containment concrete, sample data belonging to the index are collected on multiple containment concrete specimens with pipe openings of different diameters.

[0055] Step 2: Perform sample data preprocessing to obtain preprocessed measurement data;

[0056] Step 3: Construct a prediction model for the frost resistance of concrete containment structures with pipe openings based on the PSO-BP algorithm;

[0057] Step 4: Set the number of particle swarm iterations and the population size. Through calculation and iterative updates, obtain the individual optimal value and the global optimal value. Assign the global optimal value to the BP neural network.

[0058] Step 5: Based on the dataset of measured values ​​obtained in Step 2, take the number of freeze-thaw cycles and the diameter of the pipe with the opening as two influencing factors as input variables, and take the measured value of the total mass of the actual unit surface area of ​​the specimen as the output variable to train the BP neural network after the assignment in Step 4.

[0059] Step 6: Test the preprocessed sample dataset with the trained prediction model for the frost resistance of containment concrete with pipe openings based on the PSO-BP algorithm, and determine the R-value between the prediction model's prediction results and the measured dataset. 2Whether it is greater than 95%; if the R-value of the prediction model prediction result is greater than or equal to the measured value dataset. 2 If the value is less than 95%, adjust the number of particle swarm optimization iterations to obtain new optimal individual values ​​and new optimal population values, until the R-value between the prediction model's prediction and the measured dataset is found to be within acceptable limits. 2 With a success rate greater than 95%, the final prediction model is obtained. The final population size of this model is 30, and the number of iterations is 100. Using this final prediction model, the mass loss of any pipe diameter under any number of freeze-thaw cycles is predicted.

[0060] In step 1, using the mass of spalling material from a single-sided freeze-thaw cycle of the concrete with pipe openings as an index of the concrete's frost resistance means that the frost resistance of the concrete with pipe openings is evaluated using the converted data M, which is the mass of spalling material from a single-sided freeze-thaw cycle of the concrete specimen in a medium composed of 97% distilled water and 3% sodium chloride solution. The collected data is the total mass of spalling material on the specimen surface after N freeze-thaw cycles.

[0061] The conversion method is as follows:

[0062] μ s =μ f -μ b (1)

[0063] A = b 2 -πd 2 / 4 (2)

[0064]

[0065] In the formula, μ s The mass (g) of the flaking material on the specimen surface is accurate to 0.01g; μ b The mass of the filter paper (g) is accurate to 0.01g; μ f The total mass (g) of the dried filter paper and the residue from the test specimen, accurate to 0.01 g; ∑μ s Let M be the mass of the surface spalling material of the specimen after N cumulative single-sided freeze-thaw cycles, and M be the total mass of spalling material per unit test surface area of ​​a single specimen after N freeze-thaw cycles (g / m²). 2 A represents the surface area (mm²) of a single test surface of the specimen. 2 b is the side length (mm) of the test surface of the specimen; d is the diameter (mm) of the steel pipe in the concrete containment structure with pipe opening.

[0066] The pretreatment in step 2 includes: taking the arithmetic mean of the total mass of spalling material per unit test surface area of ​​multiple concrete specimens with pipe openings of the same specification in the same group as the measured value of the total mass of spalling material per unit surface area of ​​the specimens in that group.

[0067] Step 3, the construction of the prediction model for the frost resistance of containment concrete with pipe openings based on the PSO-BP algorithm, specifically includes the following steps:

[0068] Step 3-1: Determine the topology of the BP network;

[0069] Step 3-2: Generate the initial population, initializing particle velocity, position, and individual and global optimum. In step 3-2, particles update their velocity and position using the individual and global optimum, with the following update formula:

[0070]

[0071]

[0072] In the formula, V i P represents particle velocity; ω represents relation weight; d = 1, 2, ..., D, where D is the dimension of the search space; k is the current iteration number; P i For the individual best; P g For global best; X i denoted as the particle position; c1 and c2 are non-negative constants; r1 and r2 are random numbers distributed between [0,1].

[0073] Step 3-3: Calculate the error and use the calculated error as the particle fitness to update the individual best and the global best.

[0074] Steps 3-4: Update the example speed and location;

[0075] Step 3-5: Repeat steps 3-3 to 3-4 until the termination condition is met; the termination condition is that the number of iterations reaches the maximum value.

[0076] Steps 3-6: Obtain the optimized weights and thresholds of the BP network;

[0077] Steps 3-7: Calculate the error, and update the weights and thresholds;

[0078] Step 3-8: Repeat steps 3-6 to 3-7 until the termination condition is met; the termination condition is that the network prediction error is less than the given accuracy, or the number of iterations reaches the maximum value.

[0079] Steps 3-9: Output the predicted values.

[0080] Example 2:

[0081] Please refer to Figure 7 Embodiment 2 of the present invention provides a device for predicting the frost resistance of concrete containment structures with pipe openings. This device includes...

[0082] The data acquisition module is used to take the mass parameters of the spalling material after one-sided freeze-thaw of the concrete with pipe opening as the freeze-thaw resistance index of the concrete with pipe opening. Based on the selected freeze-thaw resistance index of the concrete with pipe opening, the module collects the index and obtains sample data for multiple concrete specimens with pipe opening of the same pipe diameter in each group.

[0083] The data preprocessing module is used to preprocess sample data to obtain a preprocessed dataset of measured values.

[0084] The prediction model building module is used to build a prediction model for the frost resistance of containment concrete with pipe openings based on the PSO-BP algorithm.

[0085] The prediction model training module uses the set number of particle swarm iterations and population size to obtain individual optimal values ​​and global optimal values ​​through calculation and iterative updates. The global optimal value is then assigned to the BP neural network. Based on the obtained measurement dataset, the number of freeze-thaw cycles and the diameter of the pipe with the opening are used as two influencing factors as input variables, and the total mass of the actual unit surface area of ​​the specimen is used as the output variable to train the BP neural network after the assignment in step 4.

[0086] The mass loss prediction module uses a trained prediction model based on the PSO-BP algorithm for the frost resistance of containment concrete with pipe openings to test the preprocessed sample dataset, and determines the R-value between the prediction model's prediction results and the measured dataset. 2 Whether it is greater than 95%; if the R-value of the prediction model prediction result is greater than or equal to the measured value dataset. 2 If the value is less than 95%, adjust the number of particle swarm optimization iterations to obtain new optimal individual values ​​and new optimal population values, until the R-value between the prediction model's prediction and the measured dataset is found to be within acceptable limits. 2 If the accuracy is greater than 95%, the final prediction model is obtained. Using the final prediction model, the mass loss of any pipe diameter under any number of freeze-thaw cycles can be predicted.

[0087] Example 3:

[0088] Embodiment 3 of the present invention provides a computer storage medium storing a computer program, characterized in that, when the computer program is executed by a processor, it implements the method provided in Embodiment 1 for predicting the frost resistance of containment concrete with pipe openings based on the PSO-BP algorithm.

[0089] Specific calculation and verification examples:

[0090] Please refer to Figure 5 This invention provides a method for predicting the frost resistance of concrete containment structures with pipe openings based on the PSO-BP algorithm, the specific steps of which include:

[0091] Step 1: Embed pipes with diameters of 20, 25, 32, 38, and 48 mm into cubic concrete specimens, respectively. The cubic concrete specimens are 150 mm × 150 mm × 150 mm in size and are numbered D. 20 Group, D 25 Group, D 32 Group, D 38 Group, D 48 Group D0 was used as a control, consisting of concrete without pipe openings (i.e., no pre-drilled holes). Each group of concrete specimens consisted of 3 pieces, which were cut and processed into 6 specimens measuring 150mm × 150mm × 70mm after 28 days of standard curing.

[0092] Step 2: Place the specimen in a solution of 97% distilled water and 3% sodium chloride for single-sided salt solution freeze-thaw cycles. The freeze-thaw cycle regime is as follows: Figure 3 As shown. After every four freeze-thaw cycles, the total mass M of the flaking material on the surface of the specimen was measured and pretreated, until 28 freeze-thaw cycles were completed;

[0093] Step 3: Since the final test result includes the conditions specified in GB / T50082-2009—the total mass of flaking material per unit surface area of ​​the specimen is greater than 1500 g / mm². 2 All data can be used to train the PSO-BP prediction model, with the number of freeze-thaw cycles and the diameter of the pipe with the opening as two influencing factors, and the total mass of the spalling material per unit surface area of ​​the specimen as the output variable.

[0094] Step 4: Using the trained prediction model, predict the freeze-thaw resistance of containment concrete under different freeze-thaw cycles and pipe diameters with pipe openings, and compare the predictions with the experimental results. Table 1 shows a comparison between the experimental results and the predicted results regarding the total mass of spalling material per unit surface area, where M... p For the prediction result, M e These are the results of the experiment.

[0095] Table 1 Comparison of Experimental Data and Predicted Results

[0096]

[0097]

[0098] Figure 6 This is a comparison chart of the experimental results and predicted results for M under the same pipe diameter but different freeze-thaw cycles. Figure 6 It can be seen that the predicted result of M is not much different from the experimental result, and both increase with the increase of the number of freeze-thaw cycles.

[0099] Calculate the MAPE and R of the prediction results2 As shown in Table 2.

[0100] Table 2 shows the R-values ​​of the PSO-BP prediction results. 2 and MAPE 2

[0101]

[0102] The data above shows that the PSO-BP prediction model can provide good prediction results for the freeze-thaw resistance of concrete containment structures with pipe openings, and can provide some useful guidance in practice.

[0103] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made by those skilled in the art to the technical solutions of the present invention without departing from the spirit of the present invention should fall within the protection scope defined by the present invention.

Claims

1. A method for predicting the frost resistance of concrete containment structures with pipe openings based on the PSO-BP algorithm, characterized in that: Includes the following steps: Step 1: The mass parameters of the spalling material after single-sided freeze-thaw of the concrete with pipe openings are used as the frost resistance index of the concrete with pipe openings. Based on the selected frost resistance index of the concrete with pipe openings, sample data belonging to the index are collected on multiple concrete specimens with pipe openings of different diameters. Step 2: Perform sample data preprocessing to obtain the preprocessed measurement value dataset; Step 3: Construct a prediction model for the frost resistance of concrete containment structures with pipe openings based on the PSO-BP algorithm; Step 4: Set the number of particle swarm iterations and the population size. Through calculation and iterative updates, obtain the individual optimal value and the global optimal value. Assign the global optimal value to the BP neural network. Step 5: Based on the dataset of measured values ​​obtained in Step 2, take the number of freeze-thaw cycles and the diameter of the pipe with the opening as two influencing factors as input variables, and take the measured value of the total mass of the actual unit surface area of ​​the specimen as the output variable to train the BP neural network after the assignment in Step 4. Step 6: Test the preprocessed sample dataset with the trained prediction model for the frost resistance of containment concrete with pipe openings based on the PSO-BP algorithm, and determine the R-value between the prediction model's prediction results and the measured dataset. 2 Whether it is greater than 95%; if the R-value of the prediction model prediction result is greater than or equal to the measured value dataset. 2 If the value is less than 95%, adjust the number of particle swarm optimization iterations to obtain new optimal individual values ​​and new optimal population values, until the R-value between the prediction model's prediction and the measured dataset is found to be within acceptable limits. 2 If the accuracy is greater than 95%, the final prediction model is obtained. Using the final prediction model, the mass loss of any pipe diameter under any number of freeze-thaw cycles can be predicted.

2. The method for predicting the frost resistance of concrete containment structures with pipe openings based on the PSO-BP algorithm according to claim 1, characterized in that, In step 1, using the mass of spalling material from a single-sided freeze-thaw cycle of the concrete with pipe openings as an index of the concrete's freeze-thaw resistance means that the mass of spalling material from a single-sided freeze-thaw cycle of the concrete specimen with pipe openings in a medium composed of 97% distilled water and 3% sodium chloride solution is converted and used as an index of the concrete's freeze-thaw resistance.

3. The method for predicting the frost resistance of concrete containment structures with pipe openings based on the PSO-BP algorithm according to claim 2, characterized in that, The conversion method is as follows: m s =μ f -m b (1) A=b 2 -πd 2 / 4 (2) In the formula, μ s The mass (g) of the flaking material on the specimen surface is accurate to 0.01g; μ b The mass of the filter paper (g) is accurate to 0.01g; μ f The total mass (g) of the dried filter paper and the residue from the test specimen, accurate to 0.01 g; ∑μ s Let M be the total mass of the surface spalling material of the specimen after N cumulative single-sided freeze-thaw cycles, M be the total mass of the spalling material per unit test surface area of ​​a single specimen after N freeze-thaw cycles, A be the surface area of ​​the test surface of a single specimen, b be the side length of the test surface of the specimen, and d be the diameter of the steel pipe in the concrete containment structure with pipe opening.

4. The method for predicting the frost resistance of concrete containment structures with pipe openings based on the PSO-BP algorithm according to claim 1, characterized in that, The pretreatment in step 2 includes: taking the arithmetic mean of the total mass of spalling material per unit test surface area of ​​multiple concrete specimens with pipe openings of the same specification in the same group as the measured value of the total mass of spalling material per unit surface area of ​​the specimens in that group.

5. The method for predicting the frost resistance of concrete containment structures with pipe openings based on the PSO-BP algorithm according to claim 1, characterized in that, Step 3, the construction of the prediction model for the frost resistance of containment concrete with pipe openings based on the PSO-BP algorithm, specifically includes the following steps: Step 5-1: Determine the topology of the BP network; Step 5-2: Generate the initial population, and initialize particle velocities, positions, individual optimalities, and global optimalities; Step 5-3: Calculate the error and use the calculated error as the particle fitness to update the individual best and the global best. Step 5-4: Update the example speed and location; Step 5-5: Repeat steps 5-3 to 5-4 until the termination condition is met; Steps 5-6: Assign initial weights and thresholds to the BP network based on the particle swarm positions; Steps 5-7: Calculate the error, and update the weights and thresholds; Steps 5-8 and 5-7 are repeated until the termination condition is met; Steps 5-9: Output the predicted values.

6. The method for predicting the frost resistance of concrete containment structures with pipe openings based on the PSO-BP algorithm according to claim 5, characterized in that, In step 5-3, the particle updates its velocity and position through individual and global best-case analyses, using the following update formula: In the formula, V i P represents particle velocity; ω represents relation weight; d = 1, 2, ..., D, where D is the dimension of the search space; k is the current iteration number; P i For the individual best; P g For global best; X i denoted as the particle position; c1 and c2 are non-negative constants; r1 and r2 are random numbers distributed between [0,1].

7. The method for predicting the frost resistance of concrete containment structures with pipe openings based on the PSO-BP algorithm according to claim 5, characterized in that, In step 5-5, the termination condition is that the number of iterations reaches the maximum value.

8. The method for predicting the frost resistance of concrete containment structures with pipe openings based on the PSO-BP algorithm according to claim 5, characterized in that, In steps 5-8, the termination condition is that the network prediction error is less than the given accuracy, or the number of iterations reaches the maximum value.

9. A device for predicting the frost resistance of concrete containment structures with pipe openings, characterized in that, include: The data acquisition module is used to take the mass parameters of the spalling material after one-sided freeze-thaw of the concrete with pipe opening as the freeze-thaw resistance index of the concrete with pipe opening. Based on the selected freeze-thaw resistance index of the concrete with pipe opening, the module collects the index and obtains sample data for multiple concrete specimens with pipe opening of the same pipe diameter in each group. The data preprocessing module is used to preprocess sample data to obtain a preprocessed dataset of measured values. The prediction model building module is used to build a prediction model for the frost resistance of containment concrete with pipe openings based on the PSO-BP algorithm. The prediction model training module uses the set number of particle swarm iterations and population size to obtain the individual optimal value and the global optimal value through calculation and iterative updates, and assigns the global optimal value to the BP neural network. Based on the obtained dataset of measured values, the number of freeze-thaw cycles and the diameter of the pipe with the opening are taken as two influencing factors, and the total mass of the actual unit surface area of ​​the specimen is taken as the output variable. The BP neural network after the assignment in step 4 is then trained. The mass loss prediction module uses a trained prediction model based on the PSO-BP algorithm for the frost resistance of containment concrete with pipe openings to test the preprocessed sample dataset, and determines the R-value between the prediction model's prediction results and the measured dataset. 2 Whether it is greater than 95%; if the R-value of the prediction model prediction result is greater than or equal to the measured value dataset. 2 If the value is less than 95%, adjust the number of particle swarm optimization iterations to obtain new optimal individual values ​​and new optimal population values, until the R-value between the prediction model's prediction and the measured dataset is found to be within acceptable limits. 2 If the accuracy is greater than 95%, the final prediction model is obtained. Using the final prediction model, the mass loss of any pipe diameter under any number of freeze-thaw cycles can be predicted.

10. A computer storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the method for predicting the frost resistance of containment concrete with pipe openings based on the PSO-BP algorithm as described in any one of claims 1-8.

Images