CN117113516B - Limit bearing capacity prediction method and related device for strip foundation of adjacent side slope - Google Patents
Limit bearing capacity prediction method and related device for strip foundation of adjacent side slope Download PDFInfo
- Publication number
- CN117113516B CN117113516B CN202311370770.0A CN202311370770A CN117113516B CN 117113516 B CN117113516 B CN 117113516B CN 202311370770 A CN202311370770 A CN 202311370770A CN 117113516 B CN117113516 B CN 117113516B
- Authority
- CN
- China
- Prior art keywords
- bearing capacity
- ultimate bearing
- neural network
- parameters
- convolutional neural
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 102
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 57
- 239000002689 soil Substances 0.000 claims abstract description 50
- 238000012549 training Methods 0.000 claims abstract description 46
- 238000004364 calculation method Methods 0.000 claims abstract description 43
- 238000012360 testing method Methods 0.000 claims abstract description 31
- 241000282461 Canis lupus Species 0.000 claims description 58
- 241000282421 Canidae Species 0.000 claims description 26
- 238000002790 cross-validation Methods 0.000 claims description 11
- 230000008569 process Effects 0.000 claims description 10
- 238000012795 verification Methods 0.000 claims description 10
- 238000013461 design Methods 0.000 claims description 9
- 238000005457 optimization Methods 0.000 claims description 9
- 238000004422 calculation algorithm Methods 0.000 claims description 8
- 238000011176 pooling Methods 0.000 claims description 8
- 238000004088 simulation Methods 0.000 claims description 7
- 238000000342 Monte Carlo simulation Methods 0.000 claims description 6
- 238000004458 analytical method Methods 0.000 claims description 6
- 239000006185 dispersion Substances 0.000 claims description 4
- 238000005070 sampling Methods 0.000 claims description 4
- 238000005311 autocorrelation function Methods 0.000 claims description 3
- 239000000463 material Substances 0.000 abstract description 2
- 238000004891 communication Methods 0.000 description 8
- 238000010586 diagram Methods 0.000 description 8
- 238000004590 computer program Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 230000005236 sound signal Effects 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 238000010008 shearing Methods 0.000 description 3
- 230000004913 activation Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- KLDZYURQCUYZBL-UHFFFAOYSA-N 2-[3-[(2-hydroxyphenyl)methylideneamino]propyliminomethyl]phenol Chemical compound OC1=CC=CC=C1C=NCCCN=CC1=CC=CC=C1O KLDZYURQCUYZBL-UHFFFAOYSA-N 0.000 description 1
- 238000012935 Averaging Methods 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 201000001098 delayed sleep phase syndrome Diseases 0.000 description 1
- 208000033921 delayed sleep phase type circadian rhythm sleep disease Diseases 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000002195 synergetic effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/13—Architectural design, e.g. computer-aided architectural design [CAAD] related to design of buildings, bridges, landscapes, production plants or roads
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/23—Design optimisation, verification or simulation using finite element methods [FEM] or finite difference methods [FDM]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces
Abstract
The invention provides a method and a related device for predicting the ultimate bearing capacity of a strip foundation of an adjacent side slope, and relates to the technical field of prediction of bearing capacity of the adjacent side slope, comprising the steps of obtaining foundation information of the adjacent side slope and simulating random field images of the model side slope under different soil parameters; generating a plurality of data samples using the random field image, the data set being formed from all of the data samples; building a convolutional neural network model, and training and testing the convolutional neural network model by utilizing the data set to obtain a limit bearing capacity prediction model; the method is used for solving the problem that the prediction of the existing slope foundation limit bearing capacity requires large-batch random field numerical calculation and consumes a great amount of manpower and material resources.
Description
Technical Field
The invention relates to the technical field of prediction of bearing capacity of adjacent slopes, in particular to a method and a related device for predicting limit bearing capacity of a strip foundation of an adjacent slope.
Background
Due to engineering construction requirements and environmental restrictions, foundations of bridge piers, soil retaining structures and high-rise buildings are often arranged on the tops of slopes adjacent to the slopes to form foundations adjacent to the slopes. Unlike horizontal foundations, the presence of a slope reduces the ultimate load bearing capacity of the adjacent slope foundation. The ultimate bearing capacity of the strip foundation adjacent to the side slope is mainly influenced by factors such as foundation width, slope height, slope inclination angle, soil layer cohesive force, soil layer friction angle, soil layer shear strength and the like. Soil properties such as soil layer cohesive force, friction angle, shear strength and the like are unstable along with the spatial change, so that the spatial variability of soil mass needs to be considered when analyzing the limit bearing capacity of the slope foundation. At present, the prediction of the limit bearing capacity of the slope foundation needs to be performed with a large amount of random field numerical calculation, so that a large amount of manpower and material resources are consumed, and a large amount of time cost is increased.
Disclosure of Invention
The invention aims to provide a method and a related device for predicting ultimate bearing capacity of a strip foundation adjacent to a side slope so as to solve the problems. In order to achieve the above purpose, the technical scheme adopted by the invention is as follows:
in a first aspect, the present application provides a method for predicting ultimate bearing capacity of a strip foundation adjacent to a slope, including:
acquiring basic information of adjacent slopes, constructing a model slope according to the basic information, and simulating random field images of the model slope under different soil parameters;
generating a plurality of data samples by using the random field image, calculating the ultimate bearing capacity corresponding to each data sample, forming a data set by all the data samples, and dividing the data set into a training set and a testing set;
building a convolutional neural network model, training the convolutional neural network model by using the training set, and determining the optimal super-parameters of the convolutional neural network model by using a gray wolf optimization algorithm in the training process;
substituting the optimal super parameters into a convolutional neural network model for five-fold cross validation, and then testing the convolutional neural network model by using a test set to obtain a ultimate bearing capacity prediction model;
and generating a plurality of calculation samples by using the random field image, predicting the predicted ultimate bearing capacity corresponding to each calculation sample by using the ultimate bearing capacity prediction model, and calculating the ultimate bearing capacity of the strip foundation of the adjacent side slope by using a Monte Carlo simulation method according to the predicted ultimate bearing capacities corresponding to all calculation samples.
In a second aspect, the present application further provides a device for predicting ultimate bearing capacity of a strip foundation adjacent to a slope, including:
and (3) an analog module: the method comprises the steps of obtaining basic information of an adjacent slope, constructing a model slope according to the basic information, and simulating random field images of the model slope under different soil parameters;
the data set constitutes the module: the random field image generation method comprises the steps of generating a plurality of data samples by utilizing the random field image, calculating the ultimate bearing capacity corresponding to each data sample, forming a data set by all the data samples, and dividing the data set into a training set and a testing set;
training module: the method is used for building a convolutional neural network model, training the convolutional neural network model by utilizing the training set, and determining the optimal super-parameters of the convolutional neural network model by utilizing a gray wolf optimization algorithm in the training process;
and a testing module: the method comprises the steps of substituting the optimal super parameters into a convolutional neural network model for five-fold cross validation, and then testing the convolutional neural network model by using a test set to obtain a ultimate bearing capacity prediction model;
and a prediction module: and the method is used for generating a plurality of calculation samples by utilizing the random field image, predicting the predicted ultimate bearing capacity corresponding to each calculation sample by using the ultimate bearing capacity prediction model, and calculating the ultimate bearing capacity of the adjacent slope strip foundation by utilizing a Monte Carlo simulation method according to the predicted ultimate bearing capacity corresponding to all calculation samples.
In a third aspect, the present application further provides a ultimate bearing capacity predicting apparatus adjacent to a slope strip foundation, including:
a memory for storing a computer program;
and the processor is used for realizing the ultimate bearing capacity prediction method of the adjacent slope strip foundation when executing the computer program.
In a fourth aspect, the present application further provides a readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the method for predicting ultimate bearing capacity based on a strip foundation of an adjacent side slope as described above.
The beneficial effects of the invention are as follows:
according to the invention, firstly, 200 times of random field finite element calculation is carried out, then 10-to-5 times of random field samples are rapidly calculated through a CNN network generated by training the 200 samples, and compared with the prior art, under the condition that the calculation accuracy is unchanged, the calculation efficiency is improved by about 600 times, meanwhile, the invention adopts the gray wolf optimization algorithm to optimize the super-parameters of CNN, and the problems of overlong iteration time, low convergence accuracy, easiness in search stagnation and the like in the CNN parameter optimization process are effectively solved.
Additional features and advantages of the invention will be set forth in the description which follows, and in part will be apparent from the description, or may be learned by practice of the embodiments of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims thereof as well as the appended drawings.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed in the embodiments will be briefly described below, it being understood that the following drawings only illustrate some embodiments of the present invention and therefore should not be considered as limiting the scope, and other related drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic flow chart of a method for predicting ultimate bearing capacity of a strip foundation of an adjacent side slope according to an embodiment of the invention;
FIG. 2 is a diagram showing the overall layout of a slope model according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of a convolutional neural network according to an embodiment of the present invention;
FIG. 4 is a schematic diagram of a device for predicting ultimate bearing capacity of a strip foundation of an adjacent side slope according to an embodiment of the present invention;
fig. 5 is a schematic structural diagram of a device for predicting ultimate bearing capacity of a strip foundation of an adjacent side slope according to an embodiment of the present invention.
The marks in the figure:
800. the ultimate bearing capacity prediction device is close to the slope strip foundation; 801. a processor; 802. a memory; 803. a multimedia component; 804. an I/O interface; 805. a communication component.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments of the present invention. The components of the embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations. Thus, the following detailed description of the embodiments of the invention, as presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
It should be noted that: like reference numerals and letters denote like items in the following figures, and thus once an item is defined in one figure, no further definition or explanation thereof is necessary in the following figures. Meanwhile, in the description of the present invention, the terms "first", "second", and the like are used only to distinguish the description, and are not to be construed as indicating or implying relative importance.
Example 1:
referring to fig. 1, the embodiment provides a method for predicting ultimate bearing capacity of a strip foundation of an adjacent side slope, which includes:
s1, acquiring basic information of an adjacent slope, constructing a model slope according to the basic information, and simulating random field images of the model slope under different soil parameters;
specifically, the step S1 includes:
s11, acquiring geometric information, topographic information and basic design parameters of the adjacent side slope, and constructing a model side slope by using simulation calculation software by utilizing the geometric information, topographic information and basic design parameters of the adjacent side slope;
wherein the geometric properties comprise strip base width, slope height, slope angle and the like, the base design parameters comprise base width and the like, the overall layout of the adjacent slope is shown in figure 2, in the figure,represents the ultimate bearing capacity, B represents the strip base width, < ->The slope angle is represented, and H represents the slope height;
s12, determining various soil parameters and a value range of each soil parameter;
in the embodiment, the soil cohesive force, the friction angle and the shearing strength which are uncertain along with the change of the soil space variability are taken as soil parameters (soil properties) to be considered;
s13, generating independent standard normal random variables corresponding to the soil parameters by using Latin Hypercube Sampling (LHS) according to the value range of each soil parameter；
S14, simulating a plurality of random field images under different soil parameters by using a KL series expansion method of random field image dispersion according to the independent standard normal random variables:
specifically, the method for calculating the random field is as follows:；
wherein:representing random field->Is the mean value of normal variables, +.>Standard deviation of normal variable; />Is the space point coordinate in the random field discrete domain; />The number of truncated items; />，/>The +.sup.th of the autocorrelation function respectively>The feature values and feature functions; />Is an independent standard normal random variable.
Wherein,and->Can be expressed as: />；
In the formula, the functionIs an autocorrelation function>Is random field discrete domain->，/>Is any two-point coordinate in the discrete domain.
Number of truncated itemsCan be expressed as: />；
In the method, in the process of the invention,representing the expected energy ratio factor of the random field, in order to ensure the discrete accuracy of the random field, it is generally taken that，/>、/>The horizontal and vertical lengths of the random field discrete domains, respectively.
In the embodiment, under the condition of determining the range of the cohesive force of the soil body, the method comprises the following steps ofExport 200 pieces of follow-upAirport RGB images; similarly, under the condition that friction angles and shearing strength value ranges are respectively determined, respectively deriving 200 random field RGB images to obtain 3X 200 random field images;
deleting redundant blank areas of the random field image, converting the RGB image into a gray image, converting the uint8 type (0-255) data into a double-precision type (0-1), and adjusting the image size to adapt to the size of a CNN input layer.
Based on the above embodiment, the method further includes:
s2, generating a plurality of data samples by using the random field image, calculating the ultimate bearing capacity corresponding to each data sample, and forming a data set by all the data samples according to 7:3, dividing the data set into a training set and a testing set by proportion division;
specifically, the step S2 includes: randomly combining random field images corresponding to a plurality of soil parameters to generate a plurality of data samples, wherein one data sample comprises one random field image corresponding to each soil parameter;
specifically, a random field image corresponding to the cohesive force, the friction angle and the shearing strength of the soil body is selected randomly to be combined into sample data, and 200 data samples are required to be generated in the embodiment;
and determining the ultimate bearing capacity corresponding to each data sample by adopting a finite element ultimate analysis method.
Based on the above embodiment, the method further includes:
s3, building a convolutional neural network model, training the convolutional neural network model by using the training set, and determining the optimal super-parameters of the convolutional neural network model by using a gray wolf optimization algorithm in the training process; specifically, the step S3 includes:
s31, constructing a convolutional neural network model (CNN model) by an input layer, a convolutional layer introducing a ReLU nonlinear function, an average pooling layer, a full-connection layer and an output layer;
specifically, the convolutional neural network structure comprises two ReLU+convolutional layers, two average pooling layers and one discard+full connection layer, and a CNN model structure schematic diagram is shown in figure 3;
wherein the first convolution layer usesA filter of size +.>Wherein->Is one of the super parameters to be optimized by the wolf, and the stride is 1. Since ReLU activation does not change the feature map size, the feature map size after the first convolution and activation is +.>。
Average pooling layer size ofThe stride is 2, and the characteristic diagram is reduced to 25% of the original characteristic diagram after pooling, namely。
The second convolution layer usesA filter of size +.>With a stride of 1, please refer to the architecture of VGG16 network, the number of filter is 2 times of the first convolution layer, and the size of the feature map after convolution is；
The second averaging pooling layer remains unchanged, so the size of the feature map is reduced to. The discarding layer sets the discarding rate as one of the super parameters to be optimized, the subsequent full-connection layer is similar to the shallow neural network, the neurons of two adjacent layers are in full-connection relationship, and finally the regression output is performedLayer, use MAE as loss function.
S32, taking the number of layers and the discarding rate of the full-connection layer as super parameters, and initializing the wolf group scale according to the initial range of the super parameters;
s33, obtaining the maximum iteration timesInitializing the position and constant factors of the gray wolves;
s34, calculating the fitness of each wolf individual to obtain the first three wolves with the best fitness, namely the wolves alpha, beta and delta;
s35, updating constant factors, and updating the positions of the wolves by using the updated constant factors;
specifically, the method for updating the wolf group position comprises the following steps:
in hunting, the hunting object needs to be surrounded firstly, and the distance between the wolf and the hunting object is as follows:in the method, in the process of the invention,tthe current iteration number; />And->Is a synergistic coefficient vector; />A position vector representing a current prey; />A position vector representing the current gray wolf; />Represents a constant factor +.>Linearly decreasing from 2 to 0; />And->Is [0,1 ]]Random vector of>Representing the distance between the wolf and the prey.
After the wolf surrounds the prey, the wolf alpha carries the wolf collar beta and the wolf delta to guide the wolf group to catch the prey, and the positions of the wolf alpha, the wolf beta and the wolf delta are continuously approximate to the position of the prey by utilizing the positions of the wolf alpha, the wolf beta and the wolf delta, and the updated formulas of the positions of the wolf alpha, the wolf beta and the wolf delta are as follows:in (1) the->、/>And->Respectively representing the positions of alpha, beta and delta of the wolves, < ->、/>And->Respectively represent the convergence factors of the wolf alpha, the wolf beta and the wolf delta at the time t,/I>、/>And->Indicates the positions of the wolf alpha, the wolf beta and the wolf delta at the time t,/for the wolf>、/>And->Representing the distance between the wolf alpha, the wolf beta, the wolf delta and other wolf group individuals.
When the absolute value of A is more than 1, the gray wolves are dispersed in each area as much as possible and hunting objects are searched; when |A| <1, the wolf will focus on searching for hunting in one or more areas. S36, calculating the fitness of each gray wolf after the position update so as to update the first three gray wolves with the best fitness;
if the adaptability of the position updated wolf individuals is better than the current optimal adaptability, updating the first three wolves with the best adaptability;
s37, when the update times of the wolf group positions reach the maximum iteration timesOutputting three gray wolves with the first three fitness degrees;
s38, determining super parameters by utilizing the positions of three gray wolves with the first three fitness degrees:；
representing the determination of the optimal solution for the hyper-parameters.
Based on the above embodiment, the method further includes:
s4, substituting the optimal super parameters into a convolutional neural network model for five-fold cross validation, and then testing the convolutional neural network model by using a test set to obtain a limit bearing capacity prediction model;
specifically, the step S4 includes:
s41, substituting the optimal super parameters into five CNN models;
s42, inputting the training set into five CNN models for verification to obtain five-fold cross verification models CNN-1 to CNN-5; s43, inputting the test set into a five-fold cross validation model to obtain five ultimate bearing capacity prediction results, wherein the five ultimate bearing capacity prediction results are finally represented by CNN-;
s44, calculating the average value of the five ultimate bearing capacity prediction results to obtain an ultimate bearing capacity prediction value;
s45, calculating a relative error between the predicted value of the ultimate bearing capacity and the true value of the ultimate bearing capacity in the training set, and obtaining a predicted model of the ultimate bearing capacity after training when the relative error meets the precision requirement:
in this embodiment, the relative error between the predicted value and the true value is evaluated by Root Mean Square Error (RMSE), and the calculation formula is as follows:；
wherein:for the predicted value of ultimate bearing capacity, < >>Is the true value of the ultimate bearing capacity, +.>The number of samples in the test set.
Based on the above embodiment, the method further includes:
s5, generating a plurality of calculation samples by utilizing the random field image, predicting the predicted ultimate bearing capacity corresponding to each calculation sample by utilizing the ultimate bearing capacity prediction model, and calculating the ultimate bearing capacity of the strip foundation of the adjacent side slope by utilizing a Monte Carlo simulation method according to the predicted ultimate bearing capacity corresponding to all calculation samples;
specifically, the step S5 includes:
s51, randomly combining random field images corresponding to various soil parameters to generateA number of calculation samples greater than the number of data samplesPreferably, the +.>；
S52, willSequentially inputting the calculation samples into a limit bearing capacity prediction model for calculation, and predicting to obtain a predicted limit bearing capacity corresponding to each calculation sample;
s53, calculating the average value of the predicted ultimate bearing capacities corresponding to all the calculated samples to obtain the ultimate bearing capacity of the strip foundation of the adjacent side slope:；
wherein:is the ultimate bearing capacity of the strip foundation adjacent to the side slope, < ->Is the number of monte carlo samples.
Example 2:
as shown in fig. 4, the present embodiment provides a ultimate bearing capacity predicting apparatus for a strip foundation of an adjacent side slope, the apparatus comprising:
and (3) an analog module: the method comprises the steps of obtaining basic information of an adjacent slope, constructing a model slope according to the basic information, and simulating random field images of the model slope under different soil parameters;
the data set constitutes the module: the random field image generation method comprises the steps of generating a plurality of data samples by utilizing the random field image, calculating the ultimate bearing capacity corresponding to each data sample, forming a data set by all the data samples, and dividing the data set into a training set and a testing set;
training module: the method is used for building a convolutional neural network model, training the convolutional neural network model by utilizing the training set, and determining the optimal super-parameters of the convolutional neural network model by utilizing a gray wolf optimization algorithm in the training process;
and a testing module: the method comprises the steps of substituting the optimal super parameters into a convolutional neural network model for five-fold cross validation, and then testing the convolutional neural network model by using a test set to obtain a ultimate bearing capacity prediction model;
and a prediction module: and the method is used for generating a plurality of calculation samples by utilizing the random field image, predicting the predicted ultimate bearing capacity corresponding to each calculation sample by using the ultimate bearing capacity prediction model, and calculating the ultimate bearing capacity of the adjacent slope strip foundation by utilizing a Monte Carlo simulation method according to the predicted ultimate bearing capacity corresponding to all calculation samples.
Based on the above embodiments, the simulation module includes:
an acquisition unit: the method comprises the steps of obtaining geometric information, topographic information and basic design parameters of an adjacent slope, and constructing a model slope by using simulation calculation software by utilizing the geometric information, topographic information and basic design parameters of the adjacent slope; a first determination unit: the method is used for determining various soil parameters and the value range of each soil parameter;
a first generation unit: the method comprises the steps of generating independent standard normal random variables corresponding to soil parameters by using Latin hypercube sampling according to the value range of each soil parameter;
simulation unit: the method is used for simulating a plurality of random field images under different soil parameters by adopting a KL series expansion method of random field image dispersion according to the independent standard normal random variables.
Based on the above embodiments, the data set constructing module includes:
a second generation unit: for randomly combining the random field images corresponding to the plurality of soil parameters to generate a plurality of data samples, wherein one data sample comprises one random field image corresponding to each soil parameter;
analysis unit: and the method is used for determining the ultimate bearing capacity corresponding to each data sample by adopting a finite element ultimate analysis method.
Based on the above embodiments, the training module includes:
model building unit: the method comprises the steps of constructing a convolutional neural network model by an input layer, a convolutional layer introducing a ReLU nonlinear function, an average pooling layer, a full-connection layer and an output layer;
a first initializing unit: the method comprises the steps of using the number of layers and the discarding rate of a full-connection layer as super parameters, and initializing the wolf group scale according to the initial range of the super parameters;
a second initializing unit: the method comprises the steps of initializing the position and constant factor of the wolf, wherein the position and constant factor are used for acquiring the maximum iteration times;
a first calculation unit: the method comprises the steps of calculating the fitness of each wolf individual to obtain three wolves with the first three fitness;
an updating unit: the method comprises the steps of updating constant factors, and updating the positions of the wolves by using the updated constant factors;
a second calculation unit: the method comprises the steps of calculating the fitness of each individual gray wolf after position updating to update three gray wolves with the first three fitness;
an output unit: when the update times of the wolf group position reach the maximum iteration times, outputting three gray wolves with the first three fitness degrees;
a second determination unit: for determining the hyper-parameters using the position of the three wolves of the first three fitness degrees.
Based on the above embodiments, the test module includes:
and a verification unit: and substituting the optimal super parameters into five convolutional neural network models, and inputting the training set into the five convolutional neural network models for verification to obtain a five-fold cross verification model.
A first prediction unit: the method comprises the steps of inputting a test set into a five-fold cross validation model to obtain five ultimate bearing capacity prediction results;
a third calculation unit: the method comprises the steps of calculating an average value of five ultimate bearing capacity prediction results to obtain an ultimate bearing capacity prediction value;
a fourth calculation unit: and the method is used for calculating the relative error between the predicted value of the ultimate bearing capacity and the true value of the ultimate bearing capacity in the training set, and obtaining a predicted model of the ultimate bearing capacity after training when the relative error meets the precision requirement.
Based on the above embodiments, the prediction module includes:
a combination unit: the random field images corresponding to the soil parameters are randomly combined to generate a plurality of calculation samples, and the number of the calculation samples is larger than that of the data samples;
a second prediction unit: the method comprises the steps of inputting a plurality of calculation samples into a limit bearing capacity prediction model in sequence for calculation, and predicting to obtain a predicted limit bearing capacity corresponding to each calculation sample;
a fifth calculation unit: and the average value of the predicted ultimate bearing capacities corresponding to all the calculated samples is used for calculating to obtain the ultimate bearing capacity of the strip foundation of the adjacent side slope.
It should be noted that, regarding the apparatus in the above embodiments, the specific manner in which the respective modules perform the operations has been described in detail in the embodiments regarding the method, and will not be described in detail herein.
Example 3:
corresponding to the above method embodiment, there is further provided a device for predicting the ultimate bearing capacity of a strip foundation adjacent to a slope, and the ultimate bearing capacity predicting device for a strip foundation adjacent to a slope described below and the ultimate bearing capacity predicting method for a strip foundation adjacent to a slope described above may be referred to correspondingly.
Fig. 5 is a block diagram illustrating a load-limiting capacity predicting apparatus 800 for a strip foundation adjacent to a side slope according to an exemplary embodiment. As shown in fig. 5, the ultimate bearing capacity predicting apparatus 800 for a strip foundation of an adjacent side slope may include: a processor 801, a memory 802. The ultimate bearing capacity predicting device 800 of the adjacent slope strip foundation may further include one or more of a multimedia component 803, an I/O interface 804, and a communication component 805.
The processor 801 is configured to control the overall operation of the device 800 for predicting ultimate bearing capacity of the strip foundation of the adjacent side slope, so as to complete all or part of the steps in the method for predicting ultimate bearing capacity of the strip foundation of the adjacent side slope. The memory 802 is used to store various types of data to support the operation of the ultimate capacity predicting device 800 on the adjacent slope strip foundation, which may include, for example, instructions for any application or method operating on the ultimate capacity predicting device 800 on the adjacent slope strip foundation, as well as application related data, such as contact data, messages, pictures, audio, video, and the like. The Memory 802 may be implemented by any type or combination of volatile or non-volatile Memory devices, such as static random access Memory (Static Random Access Memory, SRAM for short), electrically erasable programmable Read-Only Memory (Electrically Erasable Programmable Read-Only Memory, EEPROM for short), erasable programmable Read-Only Memory (Erasable Programmable Read-Only Memory, EPROM for short), programmable Read-Only Memory (Programmable Read-Only Memory, PROM for short), read-Only Memory (ROM for short), magnetic Memory, flash Memory, magnetic disk, or optical disk. The multimedia component 803 may include a screen and an audio component. Wherein the screen may be, for example, a touch screen, the audio component being for outputting and/or inputting audio signals. For example, the audio component may include a microphone for receiving external audio signals. The received audio signals may be further stored in the memory 802 or transmitted through the communication component 805. The audio assembly further comprises at least one speaker for outputting audio signals. The I/O interface 804 provides an interface between the processor 801 and other interface modules, which may be a keyboard, mouse, buttons, etc. These buttons may be virtual buttons or physical buttons. The communication module 805 is used for wired or wireless communication between the ultimate bearing capacity predicting device 800 of the adjacent slope strip foundation and other devices. Wireless communication, such as Wi-Fi, bluetooth, near field communication (Near FieldCommunication, NFC for short), 2G, 3G or 4G, or a combination of one or more thereof, the respective communication component 805 may thus comprise: wi-Fi module, bluetooth module, NFC module.
In an exemplary embodiment, the ultimate capacity predicting device 800 of the adjacent slope strip foundation may be implemented by one or more application specific integrated circuits (Application Specific Integrated Circuit, abbreviated as ASIC), digital signal processors (DigitalSignal Processor, abbreviated as DSP), digital signal processing devices (Digital Signal Processing Device, abbreviated as DSPD), programmable logic devices (Programmable Logic Device, abbreviated as PLD), field programmable gate arrays (Field Programmable Gate Array, abbreviated as FPGA), controllers, microcontrollers, microprocessors, or other electronic components for performing the ultimate capacity predicting method of the adjacent slope strip foundation described above.
In another exemplary embodiment, a computer readable storage medium is also provided, comprising program instructions which, when executed by a processor, implement the steps of the method of predicting ultimate bearing capacity of a proximal slope strip foundation described above. For example, the computer readable storage medium may be the memory 802 including the program instructions described above, which are executable by the processor 801 of the adjacent slope strip foundation ultimate bearing capacity prediction device 800 to perform the adjacent slope strip foundation ultimate bearing capacity prediction method described above.
Example 4:
corresponding to the above method embodiment, a readable storage medium is further provided in this embodiment, and a readable storage medium described below and a method for predicting ultimate bearing capacity of an adjacent slope strip foundation described above may be referred to correspondingly.
A readable storage medium having stored thereon a computer program which when executed by a processor performs the steps of the method for predicting ultimate bearing capacity of a strip foundation of an adjacent side slope of the above method embodiment.
The readable storage medium may be a usb disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM), a magnetic disk, or an optical disk, and the like.
The above description is only of the preferred embodiments of the present invention and is not intended to limit the present invention, but various modifications and variations can be made to the present invention by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
The foregoing is merely illustrative of the present invention, and the present invention is not limited thereto, and any person skilled in the art will readily recognize that variations or substitutions are within the scope of the present invention. Therefore, the protection scope of the invention is subject to the protection scope of the claims.
Claims (8)
1. The method for predicting the ultimate bearing capacity of the strip foundation of the adjacent side slope is characterized by comprising the following steps of:
acquiring basic information of an adjacent slope, constructing a model slope according to the basic information, and simulating random field images of the model slope under different soil parameters, wherein the method comprises the following steps:
obtaining geometric information, topographic information and basic design parameters of the adjacent side slope, and constructing a model side slope by using simulation calculation software by utilizing the geometric information, topographic information and basic design parameters of the adjacent side slope;
determining a plurality of soil parameters and a value range of each soil parameter;
according to the value range of each soil parameter, generating an independent standard normal random variable corresponding to the soil parameter by using Latin hypercube sampling;
according to the independent standard normal random variable, a KL series expansion method of random field image dispersion is adopted to simulate a plurality of random field images under different soil parameters, wherein the random field calculation method comprises the following steps:
wherein:represents the random field, μ is the mean of the normal variables, σ is the standard deviation of the normal variables; x is the space point coordinate in the random field discrete domain; m is the number of truncated items; lambda (lambda) i ，/>The i-th eigenvalue and the eigenvalue of the autocorrelation function are respectively; zeta type toy i Is an independent standard normal random variable;
generating a plurality of data samples by using the random field image, calculating the ultimate bearing capacity corresponding to each data sample, forming a data set by all the data samples, and dividing the data set into a training set and a testing set;
building a convolutional neural network model, training the convolutional neural network model by using the training set, and determining the optimal super-parameters of the convolutional neural network model by using a gray wolf optimization algorithm in the training process;
substituting the optimal super parameters into a convolutional neural network model for five-fold cross validation, and then testing the convolutional neural network model by using a test set to obtain a ultimate bearing capacity prediction model;
and generating a plurality of calculation samples by using the random field image, predicting the predicted ultimate bearing capacity corresponding to each calculation sample by using the ultimate bearing capacity prediction model, and calculating the ultimate bearing capacity of the strip foundation of the adjacent side slope by using a Monte Carlo simulation method according to the predicted ultimate bearing capacities corresponding to all calculation samples.
2. The method for predicting ultimate bearing capacity of a strip foundation of an adjacent slope according to claim 1, wherein generating a plurality of data samples using the random field image, calculating the ultimate bearing capacity corresponding to each data sample, comprises:
randomly combining random field images corresponding to a plurality of soil parameters to generate a plurality of data samples, wherein one data sample comprises one random field image corresponding to each soil parameter;
and determining the ultimate bearing capacity corresponding to each data sample by adopting a finite element ultimate analysis method.
3. The method for predicting ultimate bearing capacity of adjacent slope strip foundation according to claim 1, wherein constructing a convolutional neural network model, training the convolutional neural network model by using the training set, and determining optimal super parameters of the convolutional neural network model by using a gray wolf optimization algorithm in the training process, comprising:
constructing a convolutional neural network model by an input layer, a convolutional layer introducing a ReLU nonlinear function, an average pooling layer, a full-connection layer and an output layer;
taking the number of layers and the discarding rate of the full-connection layer as super parameters, and initializing the wolf size according to the initial range of the super parameters;
obtaining the maximum iteration times, and initializing the position and constant factor of the gray wolves;
calculating the fitness of each wolf individual to obtain three wolves with the front three fitness;
updating the constant factors, and updating the positions of the wolves by using the updated constant factors;
calculating the fitness of each gray wolf after the position update to update three gray wolves with the front three fitness;
when the update times of the wolf group position reach the maximum iteration times, outputting three gray wolves with the first three fitness degrees;
and determining the super-parameters by using the positions of three gray wolves with the first three fitness degrees.
4. The method for predicting the ultimate bearing capacity of a strip foundation of an adjacent side slope according to claim 1, wherein after substituting the optimal super parameters into a convolutional neural network model for five-fold cross validation, testing the convolutional neural network model by using a test set to obtain the ultimate bearing capacity prediction model, comprising:
substituting the optimal super parameters into five convolutional neural network models, and inputting the training set into the five convolutional neural network models for verification to obtain a five-fold cross verification model;
inputting the test set into a five-fold cross validation model to obtain five ultimate bearing capacity prediction results;
calculating the average value of the five ultimate bearing capacity prediction results to obtain an ultimate bearing capacity prediction value;
and calculating the relative error between the predicted value of the ultimate bearing capacity and the true value of the ultimate bearing capacity in the training set, and obtaining a predicted model of the ultimate bearing capacity after training when the relative error meets the precision requirement.
5. A limit bearing capacity prediction device for a strip foundation of an adjacent side slope, comprising:
and (3) an analog module: the method is used for acquiring basic information of adjacent slopes, constructing a model slope according to the basic information, simulating random field images of the model slope under different soil parameters, and comprises the following steps:
an acquisition unit: the method comprises the steps of obtaining geometric information, topographic information and basic design parameters of an adjacent slope, and constructing a model slope by using simulation calculation software by utilizing the geometric information, topographic information and basic design parameters of the adjacent slope;
a first determination unit: the method is used for determining various soil parameters and the value range of each soil parameter;
a first generation unit: the method comprises the steps of generating independent standard normal random variables corresponding to soil parameters by using Latin hypercube sampling according to the value range of each soil parameter;
simulation unit: the method is used for simulating a plurality of random field images under different soil parameters by adopting a KL series expansion method of random field image dispersion according to the independent standard normal random variables;
the data set constitutes the module: the random field image generation method comprises the steps of generating a plurality of data samples by utilizing the random field image, calculating the ultimate bearing capacity corresponding to each data sample, forming a data set by all the data samples, and dividing the data set into a training set and a testing set;
training module: the method is used for building a convolutional neural network model, training the convolutional neural network model by utilizing the training set, and determining the optimal super-parameters of the convolutional neural network model by utilizing a gray wolf optimization algorithm in the training process;
and a testing module: the method comprises the steps of substituting the optimal super parameters into a convolutional neural network model for five-fold cross validation, and then testing the convolutional neural network model by using a test set to obtain a ultimate bearing capacity prediction model;
and a prediction module: and the method is used for generating a plurality of calculation samples by utilizing the random field image, predicting the predicted ultimate bearing capacity corresponding to each calculation sample by using the ultimate bearing capacity prediction model, and calculating the ultimate bearing capacity of the adjacent slope strip foundation by utilizing a Monte Carlo simulation method according to the predicted ultimate bearing capacity corresponding to all calculation samples.
6. The ultimate bearing capacity predicting apparatus for a strip foundation of an adjacent side slope of claim 5, wherein said data set constructing module comprises:
a second generation unit: for randomly combining the random field images corresponding to the plurality of soil parameters to generate a plurality of data samples, wherein one data sample comprises one random field image corresponding to each soil parameter;
analysis unit: and the method is used for determining the ultimate bearing capacity corresponding to each data sample by adopting a finite element ultimate analysis method.
7. The ultimate bearing capacity predicting device for strip foundations on a side slope of claim 5, wherein said training module comprises:
model building unit: the method comprises the steps of constructing a convolutional neural network model by an input layer, a convolutional layer introducing a ReLU nonlinear function, an average pooling layer, a full-connection layer and an output layer;
a first initializing unit: the method comprises the steps of using the number of layers and the discarding rate of a full-connection layer as super parameters, and initializing the wolf group scale according to the initial range of the super parameters;
a second initializing unit: the method comprises the steps of initializing the position and constant factor of the wolf, wherein the position and constant factor are used for acquiring the maximum iteration times;
a first calculation unit: the method comprises the steps of calculating the fitness of each wolf individual to obtain three wolves with the first three fitness;
an updating unit: the method comprises the steps of updating constant factors, and updating the positions of the wolves by using the updated constant factors;
a second calculation unit: the method comprises the steps of calculating the fitness of each individual gray wolf after position updating to update three gray wolves with the first three fitness;
an output unit: when the update times of the wolf group position reach the maximum iteration times, outputting three gray wolves with the first three fitness degrees;
a second determination unit: for determining the hyper-parameters using the position of the three wolves of the first three fitness degrees.
8. The ultimate bearing capacity predicting device for strip foundations on a side slope of claim 5, wherein said test module comprises:
and a verification unit: the method comprises the steps of substituting optimal super parameters into five convolutional neural network models, and inputting the training set into the five convolutional neural network models for verification to obtain a five-fold cross verification model;
a first prediction unit: the method comprises the steps of inputting a test set into a five-fold cross validation model to obtain five ultimate bearing capacity prediction results;
a third calculation unit: the method comprises the steps of calculating an average value of five ultimate bearing capacity prediction results to obtain an ultimate bearing capacity prediction value;
a fourth calculation unit: and the method is used for calculating the relative error between the predicted value of the ultimate bearing capacity and the true value of the ultimate bearing capacity in the training set, and obtaining a predicted model of the ultimate bearing capacity after training when the relative error meets the precision requirement.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311370770.0A CN117113516B (en) | 2023-10-23 | 2023-10-23 | Limit bearing capacity prediction method and related device for strip foundation of adjacent side slope |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311370770.0A CN117113516B (en) | 2023-10-23 | 2023-10-23 | Limit bearing capacity prediction method and related device for strip foundation of adjacent side slope |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117113516A CN117113516A (en) | 2023-11-24 |
| CN117113516B true CN117113516B (en) | 2024-01-26 |
Family
ID=88805930
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311370770.0A Active CN117113516B (en) | 2023-10-23 | 2023-10-23 | Limit bearing capacity prediction method and related device for strip foundation of adjacent side slope |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117113516B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112100927A (en) * | 2020-09-24 | 2020-12-18 | 湖南工业大学 | Prediction method for slope deformation and soft soil foundation settlement based on GA-BP neural network |
| CN114818418A (en) * | 2022-04-11 | 2022-07-29 | 南昌工程学院 | Slope reliability analysis method based on active learning multivariate self-adaptive regression spline |
| CN115688225A (en) * | 2022-09-26 | 2023-02-03 | 吉林建筑大学 | Failure mechanism for evaluating earthquake-resistant limit bearing capacity of strip foundation close to side slope |
| CN116070675A (en) * | 2023-03-06 | 2023-05-05 | 西南交通大学 | Side slope neural network model selection method, device, equipment and readable storage medium |
| CN116070510A (en) * | 2023-01-04 | 2023-05-05 | 中国电力工程顾问集团西南电力设计院有限公司 | Slope foundation pile horizontal bearing reliability calculation method based on active learning Kriging model |
| CN116108590A (en) * | 2023-04-12 | 2023-05-12 | 西南交通大学 | Gravity type retaining wall design method, device, equipment and readable storage medium |
| CN116415334A (en) * | 2023-04-04 | 2023-07-11 | 浙江省工程勘察设计院集团有限公司 | Evaluation and analysis method for durability and stability of in-service slope |
-
2023
- 2023-10-23 CN CN202311370770.0A patent/CN117113516B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112100927A (en) * | 2020-09-24 | 2020-12-18 | 湖南工业大学 | Prediction method for slope deformation and soft soil foundation settlement based on GA-BP neural network |
| CN114818418A (en) * | 2022-04-11 | 2022-07-29 | 南昌工程学院 | Slope reliability analysis method based on active learning multivariate self-adaptive regression spline |
| CN115688225A (en) * | 2022-09-26 | 2023-02-03 | 吉林建筑大学 | Failure mechanism for evaluating earthquake-resistant limit bearing capacity of strip foundation close to side slope |
| CN116070510A (en) * | 2023-01-04 | 2023-05-05 | 中国电力工程顾问集团西南电力设计院有限公司 | Slope foundation pile horizontal bearing reliability calculation method based on active learning Kriging model |
| CN116070675A (en) * | 2023-03-06 | 2023-05-05 | 西南交通大学 | Side slope neural network model selection method, device, equipment and readable storage medium |
| CN116415334A (en) * | 2023-04-04 | 2023-07-11 | 浙江省工程勘察设计院集团有限公司 | Evaluation and analysis method for durability and stability of in-service slope |
| CN116108590A (en) * | 2023-04-12 | 2023-05-12 | 西南交通大学 | Gravity type retaining wall design method, device, equipment and readable storage medium |
Non-Patent Citations (2)
| Title |
|---|
| Slope reliability evaluation based on multi-objective grey wolf optimization-multi-kernel-based extreme learning machine agent model;Qing Ling 等;《 Bulletin of Engineering Geology and the Environment 》;2011-2024 * |
| 广西花岗岩软土地基沉降预测及承载力评价;刘先林 等;《中外公路》;第40卷(第4期);25-28 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117113516A (en) | 2023-11-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Stegen et al. | Inferring ecological processes from taxonomic, phylogenetic and functional trait β-diversity | |
| Tian et al. | Low-rise gable roof buildings pressure prediction using deep neural networks | |
| CN111598213B (en) | Network training method, data identification method, device, equipment and medium | |
| Muduli et al. | CPT-based probabilistic evaluation of seismic soil liquefaction potential using multi-gene genetic programming | |
| CN112561031A (en) | Model searching method and device based on artificial intelligence and electronic equipment | |
| CN114912578B (en) | Training method and device for structural response prediction model and computer equipment | |
| CN115048938A (en) | Statement emotion analysis method and device based on semantic and syntax dual channels | |
| CN115049176A (en) | Material performance evaluation method and device and computer equipment | |
| CN115204533A (en) | Oil-gas yield prediction method and system based on multivariable weighted combination model | |
| CN108875901B (en) | Neural network training method and universal object detection method, device and system | |
| CN114997036A (en) | Network topology reconstruction method, device and equipment based on deep learning | |
| CN117113516B (en) | Limit bearing capacity prediction method and related device for strip foundation of adjacent side slope | |
| CN116070675B (en) | Side slope neural network model selection method, device, equipment and readable storage medium | |
| CN112528428B (en) | Method and device for displaying physical parameters of engineering structure and computer equipment | |
| CN111639523B (en) | Target detection method, device, computer equipment and storage medium | |
| CN113112185A (en) | Teacher expressive force evaluation method and device and electronic equipment | |
| CN113158292A (en) | Component matching method, engineering quantity calculation method, device and electronic equipment | |
| CN108009150B (en) | Input method and device based on recurrent neural network | |
| CN116011071A (en) | Method and system for analyzing structural reliability of air building machine based on active learning | |
| CN112859034B (en) | Natural environment radar echo amplitude model classification method and device | |
| CN112581250B (en) | Model generation method, device, computer equipment and storage medium | |
| CN111475668A (en) | Picture visual reasoning method, equipment and storage medium for complex relation description | |
| Hashash et al. | Extracting inelastic metal behaviour through inverse analysis: A shift in focus from material models to material behaviour | |
| CN117291234B (en) | Method, device, system and storage medium for automatically aligning probes | |
| CN115049767B (en) | Animation editing method and device, computer equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |