Tubular pile bearing capacity prediction method based on IGWO-BP neural network

Abstract

The invention discloses a tubular pile bearing capacity prediction method based on an IGWO-BP neural network, and the method comprises the steps: 1, respectively building pile-soil finite element models considering different tubular pile sizes and soil layer physical parameters, respectively carrying out dynamic load and static load calculation, and collecting dynamic and static load test data; 2, performing feature selection on the dynamic load data obtained by calculation by using a Pearson correlation coefficient; 3, training a BP neural network model by using dynamic and static load experimental data, constructing an improved grey wolf optimization algorithm to optimize a weight and a threshold between a hidden layer and an output layer in the BP neural network, and finding an optimal weight and an optimal threshold through global search; and 4, substituting the dynamic load test data of the prediction set into the IGWO-BP network model to predict the bearing capacity of the pipe pile. According to the method, the corresponding tubular pile bearing capacity can be predicted by using the tubular pile dynamic load test data, so that high-precision prediction of the tubular pile bearing capacity can be realized, and the problems of high cost, long consumed time and the like of a tubular pile bearing capacity static load test are solved.

Description

Technical Field

[0001] The invention relates to the field of pipe pile bearing capacity testing, in particular to a pipe pile bearing capacity prediction method based on an improved IGWO-BP neural network. Background Art

[0002] Prestressed concrete pipe piles have the advantages of high single pile bearing capacity, flexible selection range, applicability to more geological terrains, controllable pile length, individual design according to different working conditions, less restrictions on on-site construction, fast construction speed and high growth rate. Compared with the same size of traditional cast-in-place piles, the single pile bearing capacity is increased by 30% to 50%; the main purpose of the pile is to transfer and bear the load applied by the superstructure, and the bearing capacity of the pile mainly depends on two aspects, namely the quality of the pile itself and the mechanical properties of the foundation soil; if the pile has problems such as insufficient bearing capacity, it may cause uneven settlement, causing major safety hazards to the later operation of the bridge, and the bearing capacity detection of this pile is particularly important; the bearing capacity detection of pipe piles is mainly divided into static load method and dynamic load method. The static load method has more accurate results, but it has the disadvantages of long time and high cost, while the dynamic load method is just the opposite. Its cost and time are lower than the static load method, but the error of the measured result is larger than the actual result, and it is currently difficult to use in actual projects; Summary of the invention

[0003] In order to overcome the shortcomings of the prior art, the present invention proposes a method for predicting the bearing capacity of pipe piles based on the IGWO-BP neural network, in order to achieve high-precision prediction of the bearing capacity of pipe piles, so as to overcome the many problems of high cost and long time consumption of static load test of pipe pile bearing capacity.

[0004] In order to achieve the above-mentioned purpose, the present invention adopts the following technical scheme:

[0005] The method for predicting the bearing capacity of a pile based on an IGWO-BP neural network is characterized in that it comprises the following steps:

[0006] Step 1: Use finite element software to establish numerical models of pipe piles and soil respectively, and obtain corresponding pipe pile finite element model and soil finite element model, thereby forming a pile-soil finite element model;

[0007] Step 2: Simulate the pile-soil behavior in real situations by changing the pile size and soil physical parameters in the pile-soil finite element model;

[0008] Step 3: Obtain dynamic and static load test data, including: pile bearing capacity characteristic vector and the displacement eigenvector of the pile top ,in, They represent the characteristic data of bearing capacity during the i-th static load test, represents the displacement characteristic data during the i-th dynamic load test;

[0009] Step 4: Calculation Each displacement eigenvalue in The Pearson correlation coefficients between them are sorted in descending order, and the top The displacement characteristic value corresponding to the Pearson correlation coefficient is obtained to obtain the displacement characteristic data after the i-th dynamic load test screening. , thus obtaining the filtered displacement feature vector ,in, express The displacement eigenvalues;

[0010] Step 5: Create a BP neural network and use As the input of the BP neural network model, As the output of the BP neural network model, the BP neural network is trained to obtain a trained pile bearing capacity prediction model;

[0011] Step 6: Use the improved gray wolf optimization algorithm to optimize the weights and thresholds of the trained pile bearing capacity prediction model to obtain the optimal weights and thresholds;

[0012] Step 7: The trained pipe pile bearing capacity measurement model corresponding to the optimal weight and the optimal threshold is used as the optimal pipe pile bearing capacity prediction model to realize the prediction of the pipe pile bearing capacity.

[0013] The characteristic of the pile bearing capacity prediction method based on IGWO-BP neural network described in the present invention is that step 2 includes the following steps:

[0014] Step 2.1: Randomly change the outer diameter, inner diameter and pile length of the pipe pile in the pipe pile finite element model within a certain range to simulate the sizes of different pipe piles in actual engineering;

[0015] Step 2.2: Randomly change the elastic modulus, cohesion and internal friction angle of each layer of soil in the soil finite element model within a certain range to simulate the soil under different geological conditions;

[0016] Step 2.3: In the pile-soil finite element model, the friction coefficient between the pile-soil contact surface is randomly changed within a certain range to obtain the pile-soil contact behavior data in actual engineering.

[0017] Furthermore, step 3 includes the following steps:

[0018] Step 3.1: Apply graded loads to the pile top to calculate the load-settlement curve of the pile under the graded loads. According to the vertical load-settlement curve, set the vertical compressive ultimate bearing capacity of the single pile to the load value corresponding to the total settlement of Δ, and set the characteristic value of the vertical compressive bearing capacity of the single pile to half of the vertical compressive ultimate bearing capacity of the single pile. Then, perform a static load test on the pile-soil finite element model to obtain the bearing capacity characteristic vector of the pipe pile. ,in, They represent the characteristic data of bearing capacity during the i-th static load test; T represents transposition; n is the total number of tests;

[0019] Step 3.2: Create a heavy hammer finite element model to apply an impact load to the pile, so as to perform a dynamic load test on the pile-soil finite element model and obtain the displacement characteristic vector of the pile top. ,in, represents the displacement characteristic data of the ith dynamic load test, and , express is the jth displacement eigenvalue in , where M is the number of displacement values.

[0020] Further, step 6 includes the following steps:

[0021] Step 6.1: Initialize the maximum number of iterations , define the current number of iterations as t, and initialize t=1;

[0022] The weights and thresholds of the trained pile bearing capacity prediction model are used as the weights and thresholds of the sth gray wolf individual in the tth generation. , thus constructing the t-th generation wolf pack of size N;

[0023] Step 6.2: Use equation (1) to calculate the number of individual gray wolves in the t-th generation wolf pack: The fitness value of , and get all the gray wolf individuals in the t-th generation wolf pack and sort them, so as to select the top three gray wolf individuals with the smallest fitness value and record them as the α wolf individuals in the t-th generation wolf pack. , β wolf individual , δ Wolf individuals ;

[0024] (1)

[0025] In formula (1), The sth gray wolf individual The corresponding pile bearing capacity prediction model is The predicted value of

[0026] Step 6.3: Use the GWO algorithm to Update and get the sth candidate gray wolf individual of the t+1th generation ;

[0027] Step 6.4: Use the DLH algorithm to Update and get the sth candidate gray wolf individual of the t+1th generation ;

[0028] Step 6.5: Calculate using formula (1) and The fitness value of the gray wolf is selected as the sth gray wolf individual of the t+1th generation. ; Thus we get the t+1 generation wolf pack;

[0029] Step 6.6: After assigning t+1 to t, return to step 6.2 and execute sequentially until So far, we get N gray wolf individuals in the wolf pack are selected, and the gray wolf individual with the smallest fitness is selected as the optimal weight and optimal threshold of the respectively trained pile bearing capacity prediction model.

[0030] Further, step 6.3 includes the following steps:

[0031] Step 6.3.1: Use formula (2) to calculate the number of individual gray wolves α wolf individuals The distance between , β wolf individual The distance between , δ Wolf individuals The distance between :

[0032] (2)

[0033] In formula (2), , , represents the 3 first coefficient vectors, and has:

[0034] (3)

[0035] In formula (3), , , represents three t-th generation random vectors in the range [0,1];

[0036] Step 6.3.2: Use equation (4) to calculate and update the sth gray wolf individual ;

[0037] (4)

[0038] In formula (4), , , Respectively represent the individual α wolf , β wolf individual , δ Wolf individuals The updated sth gray wolf individual; , , represents the 3 second coefficient vectors, and has:

[0039] (5)

[0040] In formula (5), , , represents another three t-th generation random vectors in the range [0,1]; represents the t-th generation update vector; and ;

[0041] Step 6.3.3: Calculate the sth candidate gray wolf individual of the t+1th generation using formula (6) ;

[0042] (6).

[0043] Further, step 6.4 includes the following steps:

[0044] Step 6.4.1: Calculate using formula (7) and The radius between :

[0045] (7)

[0046] In formula (7), represents the Euclidean distance;

[0047] Step 6.4.2: Use equation (8) to calculate the sth candidate gray wolf individual of the t+1th generation ;

[0048] (8)

[0049] In formula (8); , Respectively Centered at radius 2 randomly generated gray wolf individuals; Represents the tth generation of random numbers.

[0050] An electronic device of the present invention includes a memory and a processor, wherein the memory is used to store a program that supports the processor to execute the method for predicting the bearing capacity of a pipe pile, and the processor is configured to execute the program stored in the memory;

[0051] The present invention provides a computer-readable storage medium, wherein a computer program is stored on the computer-readable storage medium, and the computer program executes the steps of the method for predicting the bearing capacity of pipe piles when the computer program is run by a processor.

[0052] Compared with the prior art, the present invention has the following beneficial effects:

[0053] 1. Compared with the static load test, the present invention does not require huge funds and time, and has the advantages of high efficiency and short cycle. It overcomes the shortcomings of the static load test, such as high cost and time consumption, and terrain restrictions, thereby effectively solving the problems of long time consumption and high cost of the static load test of the pipe pile, and the difficulty of conducting the static load test of the pipe pile under certain complex terrain or soil conditions.

[0054] 2. Compared with the dynamic load test, the present invention realizes high-precision prediction of the bearing capacity of the pile by combining the dynamic load test results with the intelligent algorithm, thereby overcoming the problem of large errors in the results obtained from the dynamic load test;

[0055] 3. The model proposed in the present invention is universal. By changing the physical parameters of the soil such as the width and height of the pile and the elastic modulus of the soil, the problem of different pile sizes and soil types can be overcome, so that it can be applied to the prediction of pile bearing capacity in different regions.

[0056] 4. The present invention finds the relationship between the dynamic load displacement data and the static load bearing capacity of pipe piles through the BP neural network, so that the change of bearing capacity can be predicted by the change of dynamic load displacement of pipe piles. The initial weight and initial threshold of the BP neural network are optimized by introducing the improved gray wolf optimization algorithm, thereby effectively improving the prediction accuracy of the pipe pile bearing capacity prediction model. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 It is a specific flow chart of the method of the present invention;

[0058] Figure 2 It is a front schematic diagram of the pile-hammer-soil model of the present invention;

[0059] Figure 3 It is a schematic diagram of the size and mesh division of the finite element model of the present invention;

[0060] Figure 4 This is the flow chart of the IGWO algorithm of the present invention;

[0061] Figure 5This is a schematic diagram of the BP neural network structure of the present invention;

[0062] Figure 6 The invention is a comparison chart between the predicted value and the actual value of the bearing capacity of the pipe pile. DETAILED DESCRIPTION

[0063] In this embodiment, refer to Figure 1 A method for predicting the bearing capacity of a pipe pile based on an IGWO-BP neural network comprises the following steps:

[0064] Step 1: Use finite element software to establish numerical models of pipe piles and soil respectively, and obtain corresponding pipe pile finite element model and soil finite element model, thereby forming a pile-soil finite element model;

[0065] In this embodiment, finite element models of soil and pipe piles are established using finite element software. Figure 3 Part b and Figure 3 The pile-soil finite element model is obtained by dividing the mesh of the pile-soil finite element model into parts c. Figure 3 d part in the figure; the soil is divided into 6 layers and 7 layers. When there are 6 layers of soil, there are 3 layers of contact between the pile and the soil, and when there are 7 layers of soil, there are 4 layers of contact between the pile and the soil. In order to eliminate the influence of boundary effect, the finite element model of the soil is generally 20 times the diameter of the pile and the depth is about 2 times the depth of the pile into the soil. Therefore, the soil is built into a rectangle with a length and width of 10m, a three-layer contact of 40m in the depth direction, and a four-layer contact of 50m. In the case of three-layer contact of pile and soil, the soil is divided into five layers of soil, and the thickness of each layer from top to bottom is 5m, 5m, 12m, 9m, 9m, respectively. In the case of four-layer contact of pile and soil, the soil is divided into six layers of soil, and the thickness of each layer from top to bottom is 5m, 5m, 5m, 15m, 10m, 10m, respectively. The density of the soil is 2000 kg / m3, Poisson's ratio is 0.3, elastic modulus is 40MPa, cohesion is 20kPa, and internal friction angle is 20°; later, by changing some of the above parameters to simulate different soil conditions in actual situations, in the pile-soil finite element model, the friction coefficient on the pile-soil contact surface is taken as 0.35. Since the dynamic load test also requires an additional heavy hammer to apply impact load, an additional heavy hammer finite element model needs to be created when creating the finite element model, refer to Figure 3 Part a in the figure is a cylindrical structure with a diameter of 1.2m and a height of 0.8m. Its elastic modulus is 200GPa and its weight is 6t. It is allowed to fall freely from a distance of 70cm from the top of the pile. In the actual finite element modeling, the weight is fitted to the top of the pile and given an initial speed of 3.74m / s to achieve the same effect as free fall. The front schematic diagram of the pile-hammer-soil model is shown in Fig. Figure 2The outer diameter of the pile finite element model is 0.8m, the inner diameter is 0.6m, its length is 18m when in contact with the third layer of soil, and its length is 23m when in contact with the fourth layer of soil. The density is 2500 kg / m3, the elastic modulus is 48GPa, and the Poisson's ratio is 0.2.

[0066] Step 2: Simulate the pile-soil behavior in real situations by changing the pile size and soil physical parameters in the pile-soil finite element model;

[0067] Step 2.1: Randomly change the outer diameter, inner diameter and pile length of the pipe pile in the pipe pile finite element model within a certain range to simulate the sizes of different pipe piles in actual engineering;

[0068] In this embodiment, the pile outer diameter and pile inner diameter in the pipe pile finite element model are multiplied by their respective random numbers When the piles are in contact with the third and fourth soil layers, the pile lengths are multiplied by their respective random numbers. , the finite element models of pipe piles of different sizes are obtained to simulate the different sizes of piles in real situations.

[0069] Step 2.2: Randomly change the elastic modulus, cohesion and internal friction angle of each layer of soil in the soil finite element model within a certain range to simulate the soil under different geological conditions;

[0070] In this embodiment, the elastic modulus, cohesion, and internal friction angle of the soil in the soil finite element model are multiplied by their respective random numbers When the pile-soil three-layer contacts, multiply the thickness of the first, second, and fourth soil layers by their respective random numbers When the pile-soil contacts four layers, multiply the thickness of the first, second, third, and fifth layers of soil by their respective random numbers , the soil finite element model with different parameters is obtained to simulate the different soil conditions in real situations.

[0071] Step 2.3: In the pile-soil finite element model, the friction coefficient between the pile-soil contact surface is randomly changed within a certain range to obtain the pile-soil contact behavior data in the actual project;

[0072] In this embodiment, the pile-soil contact surface friction coefficient in the pile-soil finite element model is multiplied by its random number The pile-soil finite element models with different friction coefficients are obtained to simulate the pile-soil contact behavior under actual conditions.

[0073] Step 3: Obtain dynamic and static load test data:

[0074] By changing the pile-soil parameters, 200 groups of pile-soil four-layer contact finite element models and 100 groups of pile-soil three-layer contact finite element models were generated, and static load and dynamic load test calculations were carried out respectively.

[0075] Step 3.1: Apply graded loads on the pile top to calculate the load-settlement curve of the pile under the graded loads. According to the vertical load-settlement curve, the vertical compressive ultimate bearing capacity of the single pile is taken as the load value corresponding to the total settlement Δ=40mm, and the characteristic value of the vertical compressive bearing capacity of the single pile is taken as half of the vertical compressive ultimate bearing capacity of the single pile. The first, second and third level loads of the graded loads are 1500KN, 2000KN and 2400KN respectively, and the last level load is 7000KN. The intermediate load levels increase by 200KN. The pile-soil finite element model is subjected to a static load test to obtain the bearing capacity characteristic vector of the pipe pile. ,in, They represent the bearing capacity characteristic data of the i-th static load test respectively; T represents transposition; and n is the total number of tests.

[0076] Step 3.2: Create a heavy hammer finite element model to apply an impact load to the pile, so as to perform a dynamic load test on the pile-soil finite element model and obtain the displacement characteristic vector of the pile top. ,in, represents the displacement characteristic data of the ith dynamic load test, and , express is the jth displacement eigenvalue in , where M is the number of displacement values.

[0077] Step 4: Calculation Each displacement eigenvalue in The Pearson correlation coefficients between them are sorted in descending order, and the top The displacement characteristic value corresponding to the Pearson correlation coefficient is obtained to obtain the displacement characteristic data after the i-th dynamic load test screening. , thus obtaining the filtered displacement feature vector ,in, express The displacement eigenvalues.

[0078] Step 5: Reference Figure 5 , create a BP neural network, including: input layer, hidden layer and output layer; and As the input of the BP neural network model, As the output of the BP neural network model, the BP neural network is trained to obtain the trained pile bearing capacity prediction model.

[0079] Step 6: Reference Figure 4 , the weights and thresholds of the trained pile bearing capacity prediction model are optimized using the improved gray wolf optimization algorithm, so as to obtain the optimal weights and optimal thresholds;

[0080] Step 6.1: Initialize the maximum number of iterations , define the current number of iterations as t, and initialize t=1;

[0081] The weights and thresholds of the trained pile bearing capacity prediction model are used as the weights and thresholds of the sth gray wolf individual in the tth generation. , thus constructing the t-th generation wolf pack of size N;

[0082] Step 6.2: Use equation (1) to calculate the number of individual gray wolves in the t-th generation wolf pack: The fitness value of , and get all the gray wolf individuals in the t-th generation wolf pack and sort them, so as to select the top three gray wolf individuals with the smallest fitness value and record them as the α wolf individuals in the t-th generation wolf pack. , β wolf individual , δ Wolf individuals ;

[0083] (1)

[0084] In formula (1), The sth gray wolf individual The corresponding pile bearing capacity prediction model is The predicted value of

[0085] Step 6.3: Use the GWO algorithm to Update and get the sth candidate gray wolf individual of the t+1th generation :

[0086] Step 6.3.1: Use formula (2) to calculate the number of individual gray wolves α wolf individuals The distance between , β wolf individual The distance between , δ Wolf individuals The distance between :

[0087] (2)

[0088] In formula (2), , , represents the 3 first coefficient vectors, and has:

[0089] (3)

[0090] In formula (3), , , represents three t-th generation random vectors in the range [0,1].

[0091] Step 6.3.2: Use equation (4) to calculate and update the sth gray wolf individual ;

[0092] (4)

[0093] In formula (4), , , Respectively represent the individual α wolf , β wolf individual , δ Wolf individuals The updated sth gray wolf individual; , , represents the 3 second coefficient vectors, and has:

[0094] (5)

[0095] In formula (5), , , represents another three t-th generation random vectors in the range [0,1]; represents the t-th generation update vector; and The elements in satisfy Decreases linearly from 2 to 0.

[0096] Step 6.3.3: Calculate the sth candidate gray wolf individual of the t+1th generation using formula (6) ;

[0097] (6)

[0098] Step 6.4: Use the DLH algorithm to Update and get the sth candidate gray wolf individual of the t+1th generation :

[0099] Step 6.4.1: Calculate using formula (7) and The radius between :

[0100] (7)

[0101] In formula (7), represents the Euclidean distance.

[0102] Step 6.4.2: Use equation (8) to calculate the sth candidate gray wolf individual of the t+1th generation ;

[0103] (8)

[0104] In formula (8); , Respectively Centered at radius 2 randomly generated gray wolf individuals; Represents the tth generation of random numbers.

[0105] Step 6.5: Calculate using formula (1) and The fitness value of the gray wolf is selected as the sth gray wolf individual of the t+1th generation. ; Thus we get the t+1th generation wolf pack.

[0106] Step 6.6: After assigning t+1 to t, return to step 6.2 and execute sequentially until So far, we get N gray wolf individuals in the wolf pack are selected, and the gray wolf individual with the smallest fitness is selected as the optimal weight and optimal threshold of the pile bearing capacity prediction model after training respectively;

[0107] Step 7: The trained pile bearing capacity measurement model corresponding to the optimal weight and the optimal threshold is used as the optimal pile bearing capacity prediction model to realize the prediction of pile bearing capacity. The number of test set samples is 20. The comparison between the predicted values ​​and the true values ​​of the 20 groups of pile bearing capacities in the final test set is shown in the figure below. Figure 6 As shown, the prediction results are more accurate.

[0108] In this embodiment, an electronic device includes a memory and a processor, wherein the memory is used to store a program that supports the processor to execute the above method, and the processor is configured to execute the program stored in the memory.

[0109] In this embodiment, a computer-readable storage medium stores a computer program on the computer-readable storage medium, and the computer program executes the steps of the above method when executed by a processor.

Claims

1. A method for predicting the bearing capacity of pipe piles based on IGWO-BP neural network, characterized in that: The steps include: Step 1: Use finite element software to establish numerical models of pipe piles and soil respectively, and obtain corresponding pipe pile finite element model and soil finite element model, thereby forming a pile-soil finite element model; Step 2: Simulate the pile-soil behavior in real situations by changing the pile size and soil physical parameters in the pile-soil finite element model; Step 3: Obtain dynamic and static load test data, including: pile bearing capacity characteristic vector and the displacement eigenvector of the pile top ,in, They represent the characteristic data of bearing capacity during the i-th static load test, represents the displacement characteristic data during the i-th dynamic load test; Step 4: Calculation Each displacement eigenvalue in The Pearson correlation coefficients between them are sorted in descending order, and the top The displacement characteristic value corresponding to the Pearson correlation coefficient is obtained to obtain the displacement characteristic data after the i-th dynamic load test screening. , thus obtaining the filtered displacement feature vector ,in, express The displacement eigenvalues; Step 5: Create a BP neural network and use As the input of the BP neural network model, As the output of the BP neural network model, the BP neural network is trained to obtain a trained pile bearing capacity prediction model; Step 6: Use the improved gray wolf optimization algorithm to optimize the weights and thresholds of the trained pile bearing capacity prediction model to obtain the optimal weights and thresholds; Step 7: The trained pipe pile bearing capacity measurement model corresponding to the optimal weight and the optimal threshold is used as the optimal pipe pile bearing capacity prediction model to realize the prediction of the pipe pile bearing capacity.

2. The method for predicting the bearing capacity of pipe piles based on IGWO-BP neural network according to claim 1, characterized in that: Step 2 includes the following steps: Step 2.1: Randomly change the outer diameter, inner diameter and pile length of the pipe pile in the pipe pile finite element model within a certain range to simulate the sizes of different pipe piles in actual engineering; Step 2.2: Randomly change the elastic modulus, cohesion and internal friction angle of each layer of soil in the soil finite element model within a certain range to simulate the soil under different geological conditions; Step 2.3: In the pile-soil finite element model, the friction coefficient between the pile-soil contact surface is randomly changed within a certain range to obtain the pile-soil contact behavior data in actual engineering.

3. The method for predicting the bearing capacity of pipe piles based on IGWO-BP neural network according to claim 2 is characterized in that: Step 3 includes the following steps: Step 3.1: Apply graded loads to the pile top to calculate the load-settlement curve of the pile under the graded loads. According to the vertical load-settlement curve, set the vertical compressive ultimate bearing capacity of the single pile to the load value corresponding to the total settlement of Δ, and set the characteristic value of the vertical compressive bearing capacity of the single pile to half of the vertical compressive ultimate bearing capacity of the single pile. Then, perform a static load test on the pile-soil finite element model to obtain the bearing capacity characteristic vector of the pipe pile. ,in, They represent the characteristic data of bearing capacity during the i-th static load test; T represents transposition; n is the total number of tests; Step 3.2: Create a heavy hammer finite element model to apply an impact load to the pile, so as to perform a dynamic load test on the pile-soil finite element model and obtain the displacement characteristic vector of the pile top. ,in, represents the displacement characteristic data of the ith dynamic load test, and , express is the jth displacement eigenvalue in , where M is the number of displacement values.

4. The method for predicting the bearing capacity of pipe piles based on IGWO-BP neural network according to claim 3 is characterized in that: Step 6 includes the following steps: Step 6.1: Initialize the maximum number of iterations , define the current number of iterations as t, and initialize t=1; The weights and thresholds of the trained pile bearing capacity prediction model are used as the weights and thresholds of the sth gray wolf individual in the tth generation. , thus constructing the t-th generation wolf pack of size N; Step 6.2: Use equation (1) to calculate the number of individual gray wolves in the t-th generation wolf pack: The fitness value of , and get all the gray wolf individuals in the t-th generation wolf pack and sort them, so as to select the top three gray wolf individuals with the smallest fitness value and record them as the α wolf individuals in the t-th generation wolf pack. , β wolf individual , δ Wolf individuals ; (1) In formula (1), The sth gray wolf individual The corresponding pile bearing capacity prediction model is The predicted value of Step 6.3: Use the GWO algorithm to Update and get the sth candidate gray wolf individual of the t+1th generation ; Step 6.4: Use the DLH algorithm to Update and get the sth candidate gray wolf individual of the t+1th generation ; Step 6.5: Calculate using formula (1) and The fitness value of the gray wolf is selected as the sth gray wolf individual of the t+1th generation. ; Thus we get the t+1 generation wolf pack; Step 6.6: After assigning t+1 to t, return to step 6.2 and execute sequentially until So far, we get N gray wolf individuals in the wolf pack are selected, and the gray wolf individual with the smallest fitness is selected as the optimal weight and optimal threshold of the respectively trained pile bearing capacity prediction model.

5. The method for predicting the bearing capacity of pipe piles based on IGWO-BP neural network according to claim 4 is characterized in that: Step 6.3 includes the following steps: Step 6.3.1: Use formula (2) to calculate the number of individual gray wolves α wolf individuals The distance between , β wolf individual The distance between , δ Wolf individuals The distance between : (2) In formula (2), , , represents the 3 first coefficient vectors, and has: (3) In formula (3), , , represents three t-th generation random vectors in the range [0,1]; Step 6.3.2: Use equation (4) to calculate and update the sth gray wolf individual ; (4) In formula (4), , , Respectively represent the individual α wolf , β wolf individual , δ Wolf individuals The updated sth gray wolf individual; , , represents the 3 second coefficient vectors, and has: (5) In formula (5), , , represents another three t-th generation random vectors in the range [0,1]; represents the t-th generation update vector; and ; Step 6.3.3: Calculate the sth candidate gray wolf individual of the t+1th generation using formula (6) ; (6)。 6. The method for predicting the bearing capacity of pipe piles based on IGWO-BP neural network according to claim 5, characterized in that: Step 6.4 includes the following steps: Step 6.4.1: Calculate using formula (7) and The radius between : (7) In formula (7), represents the Euclidean distance; Step 6.4.2: Use equation (8) to calculate the sth candidate gray wolf individual of the t+1th generation ; (8) In formula (8); , Respectively Centered at radius 2 randomly generated gray wolf individuals; Represents the tth generation of random numbers.

7. An electronic device, comprising a memory and a processor, characterized in that: The memory is used to store a program that supports the processor to execute the method for predicting the bearing capacity of a pipe pile as described in any one of claims 1 to 6, and the processor is configured to execute the program stored in the memory.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for predicting the bearing capacity of a pipe pile according to any one of claims 1 to 6 are executed.

Images