CN111582632A - Multi-factor safety stage prediction method for whole process of underground large space construction - Google Patents

Multi-factor safety stage prediction method for whole process of underground large space construction Download PDF

Info

Publication number
CN111582632A
CN111582632A CN202010224430.7A CN202010224430A CN111582632A CN 111582632 A CN111582632 A CN 111582632A CN 202010224430 A CN202010224430 A CN 202010224430A CN 111582632 A CN111582632 A CN 111582632A
Authority
CN
China
Prior art keywords
construction
neural network
layer
safety prediction
network model
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.)
Granted
Application number
CN202010224430.7A
Other languages
Chinese (zh)
Other versions
CN111582632B (en
Inventor
肖清华
雷升祥
王立新
何亚涛
李聪明
李储军
汪珂
韩翔宇
熊强
邱泽民
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Southwest Jiaotong University
China Railway First Survey and Design Institute Group Ltd
Original Assignee
Southwest Jiaotong University
China Railway First Survey and Design Institute Group Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Southwest Jiaotong University, China Railway First Survey and Design Institute Group Ltd filed Critical Southwest Jiaotong University
Priority to CN202010224430.7A priority Critical patent/CN111582632B/en
Publication of CN111582632A publication Critical patent/CN111582632A/en
Application granted granted Critical
Publication of CN111582632B publication Critical patent/CN111582632B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
    • G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
    • 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
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/08—Construction

Abstract

The invention discloses a multi-factor safety prediction method and a system for the whole process of underground large-space construction, wherein the method comprises the following steps: constructing and training a safety prediction neural network model before construction so as to form a nonlinear mapping relation from an input layer to an output layer; constructing and training a safety prediction neural network model in construction so as to form a nonlinear mapping relation from an input layer to an output layer; and connecting the pre-construction safety prediction neural network model and the in-construction safety prediction neural network model in series to form a construction prediction series model, and performing staged safety prediction in the whole construction process by using the construction prediction series model. The construction safety prediction model is established based on the neural network with the characteristics of self-adaptability, nonlinearity and strong fault tolerance, the safety prediction can be carried out on the underground large-space construction before the construction without depending on the internal working mechanism of a rock-soil system, the real-time prediction is carried out in the construction, and the series model is formed for carrying out the whole construction process prediction.

Description

Multi-factor safety stage prediction method for whole process of underground large space construction
Technical Field
The invention relates to the technical field of civil engineering construction, in particular to a multi-factor safety prediction method and system for the whole process of underground large-space construction.
Background
Underground engineering practice shows that geological environment, hydrological environment, surrounding building environment and construction method in the construction process all influence deformation, stress, strain, settlement, displacement and the like in construction, and in most cases, engineering technicians can only master the changes of deformation, stress, strain, settlement, displacement and the like by collecting monitoring data in construction, so as to improve the construction method, and cannot predict the changes roughly before construction, which often causes various disasters and construction safety accidents due to untimely improvement.
In addition, a theoretical calculation method is mainly adopted to predict multiple safety factors such as deformation, stress, strain, settlement, displacement and the like in construction at present. However, due to lack of cognition on the internal working mechanism of the geotechnical system, great difficulty is inevitably existed in establishing a corresponding theoretical calculation expression, and the prediction calculation of the safety interference factor related to the geotechnical engineering system shows very complicated high-order nonlinear characteristics, while the nonlinear calculation has certain difficulty.
The neural network is particularly suitable for processing various nonlinear problems due to the characteristics of adaptability, nonlinearity, strong fault tolerance and the like. It can extract the causal relationships implied in the samples through the learning of a large number of samples. Therefore, the neural network provides a research idea completely different from mathematical modeling for the field of underground engineering, avoids a complex constitutive model, and becomes an effective way for solving the problem of underground engineering. Meanwhile, the neural network has a lot of applications in other fields, for example, an evaluation method of the rainstorm disaster risk of the foundation side slope of the section power transmission line tower is used for establishing a mapping relation between the control factors and the rainstorm landslide accident rate by using disaster statistics and artificial rainfall side slope erosion test results and combining with an improved hierarchical analysis calculation program to obtain the evaluation result of the rainstorm disaster of the side slope of each section of the line. The method comprises two stages before construction and in construction, the whole construction process is considered through establishing a series model of the two stages, safety prediction is carried out before construction and real-time prediction is carried out in construction, safety judgment is carried out through relevant specifications or a grading system, and the safety of the whole process construction is guaranteed.
Disclosure of Invention
At least one of the objectives of the present invention is to overcome the above problems in the prior art, and to provide a multi-factor safety prediction method and system for underground large-space construction, which is based on a neural network with strong adaptivity, nonlinearity, and fault tolerance, and can perform safety prediction on the underground large-space construction before construction and perform real-time prediction during construction without depending on the internal working mechanism of the geotechnical system.
In order to achieve the above object, the present invention adopts the following aspects.
A multi-factor safety prediction method for the whole process of underground large space construction comprises the following steps:
step 1, constructing a safety prediction neural network model before construction, and training the safety prediction neural network model before construction by using a first training sample so as to enable the performance of the safety prediction neural network model before construction to tend to be stable and form a nonlinear mapping relation from an input layer to an output layer;
wherein, the input parameters of the input layer of the safety prediction neural network before construction are as follows: engineering geological conditions, hydrological conditions, surrounding building environment, construction method, management level and construction level; the output parameters of the prediction neural network before construction are the predicted values of corresponding stress, strain, displacement and settlement;
step 2, constructing a safety prediction neural network model in construction, and training the safety prediction neural network model in construction by using a second training sample so as to enable the performance of the safety prediction neural network in construction to tend to be stable before construction and form a nonlinear mapping relation from an input layer to an output layer of the safety prediction neural network in construction;
wherein, the input parameters of the input layer of the safety prediction neural network model in the construction are as follows: input values of stress, strain, displacement and settlement of a certain time node in construction; the output parameters of the safety prediction neural network model in construction are predicted values of stress, strain, displacement and settlement of the next time node in construction;
and 3, connecting the pre-construction safety prediction neural network model and the in-construction safety prediction neural network model in series by taking the output parameters of the pre-construction safety prediction neural network model as the input parameters of the in-construction safety prediction neural network model to form a construction prediction series model, and performing staged safety prediction in the whole construction process by using the construction prediction series model.
Preferably, the method further comprises: and inputting the stress, strain, displacement and settlement monitored in real time in the construction into the in-construction safety prediction neural network model so as to predict the stress, strain, displacement and settlement of the next time node in real time in the construction through the in-construction safety prediction neural network model.
Preferably, the current output value of the node weight of each layer of the construction prediction series model is calculated by the following formula:
Figure BDA0002427175250000031
wherein the content of the first and second substances,
Figure BDA0002427175250000032
is the input sum of the ith element of the kth layer;
Figure BDA0002427175250000033
is the output of the ith element of the kth layer;
Figure BDA0002427175250000034
the connection weight value from the ith element of the kth-1 layer to the jth element of the kth layer is obtained; f is an excitation function;
Figure BDA0002427175250000035
is the actual output of the jth element of the mth layer associated with the weight vector W and the input vector X, where the mth layer is the output layer.
Preferably, the weight of each layer of node of the construction prediction series model is adjusted by the following formula:
Figure BDA0002427175250000036
when k is equal to m, the number of the symbols is m,
Figure BDA0002427175250000037
when k is<When m is greater than the total number of the carbon atoms,
Figure BDA0002427175250000038
wherein the content of the first and second substances,
Figure BDA0002427175250000039
is the input sum of the ith element of the kth layer;
Figure BDA00024271752500000310
is the output of the ith element of the kth layer;
Figure BDA00024271752500000311
the connection weight value from the ith element of the kth-1 layer to the jth element of the kth layer is obtained; f is an excitation function, which may be Sc(x)；
Figure BDA0002427175250000041
Is the actual output of the jth element of the mth layer relative to the weight vector W and the input vector X; y isiIs the desired output of the jth element of the mth layer associated with the weight vector W and the input vector X, wherein the mth layer is the output layer.
Preferably, when the pre-construction safety prediction neural network model and the in-construction safety prediction neural network model are constructed, the maximum number of iterations is set to 5000, and the learning rate η is set to 0.5.
Preferably, when the error rate of the pre-construction safety prediction neural network model and the error rate of the in-construction safety prediction neural network model are smaller than a preset value, the performance of the pre-construction safety prediction neural network tends to be stable.
In a further embodiment of the invention, the invention also provides a multi-factor safety prediction system for the whole process of underground large space construction, which comprises at least one processor and a memory which is in communication connection with the at least one processor; the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the above-described method.
In summary, due to the adoption of the technical scheme, the invention at least has the following beneficial effects:
establishing a safety prediction neural network model before construction and a safety prediction neural network model during construction based on a neural network with strong adaptivity, nonlinearity and fault tolerance, taking engineering geological conditions, hydrological conditions, surrounding building environment, construction method, management level and construction level as input variables of the safety prediction neural network model before construction, taking values of stress, strain, displacement and settlement corresponding to the input variables as output vectors of the safety prediction neural network model before construction, performing safety prediction on large underground space construction before construction, connecting the safety prediction neural network models during construction in series to form a construction prediction series model, and performing staged safety prediction of the whole construction process by using the construction prediction series model; the prediction method comprehensively considers two stages before construction and in construction, considers the whole construction process by establishing a series model of the two stages, carries out safety prediction before construction and real-time prediction in construction, and carries out safety judgment through construction specifications and/or a grading system so as to ensure the safety of the whole construction process.
And inputting the stress, strain, displacement and settlement monitored in real time in the construction into the in-construction safety prediction neural network model so as to predict the stress, strain, displacement and settlement of the next time node in real time in the construction through the in-construction safety prediction neural network model.
Drawings
FIG. 1 is a flow diagram of a pre/mid construction safety prediction neural network model training process, according to an exemplary embodiment of the present invention.
FIG. 2 is a pre-construction safety prediction neural network model topology diagram in accordance with an exemplary embodiment of the present invention.
FIG. 3 is a diagram of a safety-in-construction predictive neural network model topology, according to an exemplary embodiment of the invention.
FIG. 4 is a block diagram of construction prediction tandem model construction staging prediction according to an exemplary embodiment of the present invention.
FIG. 5 is a schematic structural diagram of an overall process multi-factor safety prediction system for underground large space construction according to an exemplary embodiment of the invention.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and embodiments, so that the objects, technical solutions and advantages of the present invention will be more clearly understood. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Example 1
The embodiment provides a multi-factor safety prediction method for the whole process of underground large space construction, which comprises the following steps:
step 1, constructing a safety prediction neural network model before construction, and training the safety prediction neural network model before construction by using a first training sample so as to enable the performance of the safety prediction neural network model before construction to tend to be stable and form a nonlinear mapping relation from an input layer to an output layer; and normalizing the data of the stress, strain, displacement and settlement corresponding to the data under different engineering geological conditions, hydrological conditions, surrounding building environments, construction methods, management levels and construction levels to obtain a first training sample.
And (4) determining input parameters. The method comprises the steps of firstly considering all projects which influence the safety and the environmental safety of the engineering during construction, such as geological conditions, hydrological conditions, surrounding building environments, construction methods and the like, and then dividing the projects according to the reference standard, the standard and the expert demonstration and determining the numerical value of the input parameter through normalization.
Determining an output parameter, based on the measured data, normalizing the output parameter; and establishing a database, namely establishing the database by using the engineering conditions, namely the input parameters, and establishing the database by using the measured data, namely the output parameters.
And (4) operation of the neural network. The learning algorithm is a multilayer neural network algorithm guided by teacher signals, is a supervised learning process, and is used for learning according to given (input and output) sample pairs and reflecting the learning effect by adjusting network connection weights.
Neural networks have two states. In the learning stage, the inputs of the learning sample pairs are added to the input end of the network, and the outputs are generated in the input and excitation function (Sigmoid function) mode in each layer of neurons along the forward direction (i.e. the input layer-the output layer). Then, the difference between the actual output value and the expected output value of the output layer neuron is reversely propagated to each layer of neuron, and each connection weight value is correspondingly adjusted according to the size and the sign of the error. This process continues until the neural network weight connection can produce a given output result with some accuracy given the input sample conditions, i.e., the learning phase is considered to be complete. In the working stage, when the sample to be tested is input into the input end of the learnt neural network, the neural network generates the required solution at the output end in an interpolation or extension mode according to the principle of similar output.
Specifically, as shown in fig. 1, the implementation steps are reading in a learning sample, normalizing data (normalization in this patent), initializing a weight of a neural network, calculating an output value of a hidden node, calculating an output value of an output node, calculating an error of an output layer, calculating an error of a hidden node, adjusting a weight, ending if the error is within an allowable error range, and performing training by normalizing again if the error is not within the allowable error range.
The process of constructing the pre-construction safety prediction neural network model shown in fig. 2 includes: and determining initial parameters. Weighted values w of the networkijThe number of hidden layers and hidden units of the network is determined by different specific problems.
The number of the hidden layer units is more, and the calculation formula of the number of the hidden layer units is more, and the formula is as follows:
n1＝log2n (1)
in the formula: n is1-the number of the hidden layer units,
n-number of input layer units.
And the number of samples, namely, the logarithm of the samples has a close relation with the number of the hidden layers and the number of the units of the hidden layers during learning. The more the hidden layers are, the more accurate the learned connection weights are, but the worse the network generalization capability is. According to research, the relation between the adjustable total number of connection weights W of the network and the necessary logarithm N of the training sample can be approximately expressed to enable the multi-layer network to have generalization capability
Figure BDA0002427175250000071
τ -coefficient, taken to be about 10.
At the time of learning, a condition for termination of learning is given. Two methods are generally used for network learning termination: first, a given error minimum value is set, and the method is terminated when the actual output error is smaller than the given error; the second is to specify the number of iterations (e.g., 5000). The former is adopted in the present system. In this embodiment, the maximum number of iterations is set to 5000, and the learning rate η is set to 0.5.
And (5) training a neural network model. The first of the P (input, output) sample pairs is input.
The actual output value is calculated. According to the formula
Figure BDA0002427175250000081
And calculating the actual output value of each layer element of the network. Adjusting the weight of each connection according to the formula
Figure BDA0002427175250000082
Adjusting each connection weight, wherein:
when k is equal to m, the number of the symbols is m,
Figure BDA0002427175250000083
when k is<When m is greater than the total number of the carbon atoms,
Figure BDA0002427175250000084
and after the first sample pair is finished, inputting a subsequent sample pair, and repeating the steps till the end. Recycle P sample pairs until wijAnd tends to be stable until unchanged. And when the network training satisfies that r is less than or equal to r, the learning process is ended.
The symbolic meanings in the above calculation formula are as follows:
Figure BDA0002427175250000085
the input sum of the ith element of the kth layer;
Figure BDA0002427175250000086
the output of the ith element of the kth layer;
Figure BDA0002427175250000087
the connection weight of the ith element of the k-1 th layer to the jth element of the kth layer;
f: the excitation function, which may be Sc(x)；
Figure BDA0002427175250000088
The actual output of the jth element of the mth layer (i.e., the output layer) relative to the weight vector W and the input vector X;
yi: the desired output of the jth element of the mth layer (i.e., the output layer) in relation to the weight vector W and the input vector X.
Step 2, constructing a safety prediction neural network model in construction as shown in fig. 3, and training the safety prediction neural network model in construction by using a second training sample so as to enable the performance of the safety prediction neural network in construction to tend to be stable before construction and form a nonlinear mapping relation from an input layer to an output layer of the safety prediction neural network in construction; and the stress, strain, displacement and settlement data of a certain time node in the construction and the stress, strain, displacement and settlement data of the next time node are normalized to form a second training sample.
Similar to step 1, the determination of the parameters is entered. Prediction in construction: the input parameters of the input layer of the safety prediction neural network model in the construction are as follows: input values of stress, strain, displacement and settlement of a certain time node in construction; the output parameter of the safety prediction neural network model in the construction is the next one in the constructionPredicting the stress, strain, displacement and settlement of the time node; and determining initial parameters. Weighted values w of the networkijGiving a small non-zero random real number initial value, setting a learning rate η and an inertia coefficient α, wherein the random initial weight values are different, and the final weight values are also different, the number of hidden layers and hidden units of the network are determined by different specific problems, a specific determination method comprises the steps of inputting P (stress, strain, displacement and settlement data of a certain time node in construction, stress, strain, displacement and settlement data of a next time node) first sample pairs according to the description and the formulas (1) and (2), calculating an actual output value, inputting subsequent sample pairs after adjusting the first sample pairs of all connection weight values according to the formulas (3) and the formulas (4) to (6), repeating the steps until the steps are finished, and circularly utilizing the P sample pairs until wijAnd tends to be stable until unchanged. And when the network training satisfies that r is less than or equal to r, the learning process is ended. After the model training is finished, the working stage can be carried out, the safety classification can be carried out, and the corresponding emergency and coping method can be determined according to the classification.
And 3, connecting the pre-construction safety prediction neural network model and the in-construction safety prediction neural network model in series by taking the output parameters of the pre-construction safety prediction neural network model as the input parameters of the in-construction safety prediction neural network model to form a construction prediction series model, and performing staged safety prediction in the whole construction process by using the construction prediction series model.
As shown in fig. 2, a prediction flow of the predictive series model is constructed. In actual application, two safety prediction neural network models before and during construction are used in series. The series model firstly inputs engineering geology, hydrogeology, surrounding environment conditions, a construction method, a management level and a construction level before construction, predicts stress, strain, displacement and settlement before a certain construction process in construction through a prediction model before construction, predicts the stress, strain, displacement and settlement of monitoring data at the next construction time point in construction through a prediction model in construction, and simultaneously predicts the stress, strain, displacement and settlement of the monitoring data at the next construction time point in construction through the prediction model in construction according to the stress, strain, displacement and settlement of the monitoring data in construction, so as to predict the stress, strain, displacement and settlement of the monitoring data at the next construction time point in construction in real time.
Furthermore, when the method is actually used, data of stress, strain, displacement and settlement of a current time node monitored in real time in the construction process are input into the in-construction safety prediction neural network model, so that the stress, strain, displacement and settlement of the next time node can be predicted in real time in construction through the in-construction safety prediction neural network model.
Example 2
Fig. 5 illustrates a multi-factor safety prediction system for the whole process of underground large space construction according to an exemplary embodiment of the present invention, namely, an electronic device 310 (e.g., a computer server with program execution function) including at least one processor 311, a power supply 314, and a memory 312 and an input/output interface 313 communicatively connected to the at least one processor 311; the memory 312 stores instructions executable by the at least one processor 311, the instructions being executable by the at least one processor 311 to enable the at least one processor 311 to perform a method disclosed in any one of the embodiments; the input/output interface 313 may include a display, a keyboard, a mouse, and a USB interface for inputting/outputting data; the power supply 314 is used to provide power to the electronic device 310.
Those skilled in the art will understand that: all or part of the steps for realizing the method embodiments can be completed by hardware related to program instructions, the program can be stored in a computer readable storage medium, and the program executes the steps comprising the method embodiments when executed; and the aforementioned storage medium includes: various media that can store program codes, such as a removable Memory device, a Read Only Memory (ROM), a magnetic disk, or an optical disk.
When the integrated unit of the present invention is implemented in the form of a software functional unit and sold or used as a separate product, it may also be stored in a computer-readable storage medium. Based on such understanding, the technical solutions of the embodiments of the present invention may be essentially implemented or a part contributing to the prior art may be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the methods described in the embodiments of the present invention. And the aforementioned storage medium includes: a removable storage device, a ROM, a magnetic or optical disk, or other various media that can store program code.
The foregoing is merely a detailed description of specific embodiments of the invention and is not intended to limit the invention. Various alterations, modifications and improvements will occur to those skilled in the art without departing from the spirit and scope of the invention.

Claims (7)

1. A multi-factor safety prediction method for the whole process of underground large space construction is characterized by comprising the following steps:
step 1, constructing a safety prediction neural network model before construction, and training the safety prediction neural network model before construction by using a first training sample so as to enable the performance of the safety prediction neural network model before construction to tend to be stable and form a nonlinear mapping relation from an input layer to an output layer;
wherein, the input parameters of the input layer of the safety prediction neural network before construction are as follows: engineering geological conditions, hydrological conditions, surrounding building environment, construction method, management level and construction level; the output parameters of the prediction neural network before construction are the predicted values of corresponding stress, strain, displacement and settlement;
step 2, constructing a safety prediction neural network model in construction, and training the safety prediction neural network model in construction by using a second training sample so as to enable the performance of the safety prediction neural network in construction to tend to be stable before construction and form a nonlinear mapping relation from an input layer to an output layer of the safety prediction neural network in construction;
wherein, the input parameters of the input layer of the safety prediction neural network model in the construction are as follows: input values of stress, strain, displacement and settlement of a certain time node in construction; the output parameters of the safety prediction neural network model in construction are predicted values of stress, strain, displacement and settlement of the next time node in construction;
and 3, connecting the pre-construction safety prediction neural network model and the in-construction safety prediction neural network model in series by taking the output parameters of the pre-construction safety prediction neural network model as the input parameters of the in-construction safety prediction neural network model to form a construction prediction series model, and performing staged safety prediction in the whole construction process by using the construction prediction series model.
2. The method of claim 1, further comprising: and inputting the stress, strain, displacement and settlement monitored in real time in the construction into the in-construction safety prediction neural network model so as to predict the stress, strain, displacement and settlement of the next time node in real time in the construction through the in-construction safety prediction neural network model.
3. The method of claim 1, wherein the current output value of the node weight of each layer of the construction prediction series model is calculated by the following formula:
Figure FDA0002427175240000021
wherein the content of the first and second substances,
Figure FDA0002427175240000022
is the input sum of the ith element of the kth layer;
Figure FDA0002427175240000023
is the output of the ith element of the kth layer;
Figure FDA0002427175240000024
the connection weight value from the ith element of the kth-1 layer to the jth element of the kth layer is obtained; f is an excitation function;
Figure FDA0002427175240000025
is the actual output of the jth element of the mth layer associated with the weight vector W and the input vector X, where the mth layer is the output layer.
4. The method of claim 3, wherein the weight of each layer node of the construction prediction series model is adjusted by the following formula:
Figure FDA0002427175240000026
when k is equal to m, the number of the symbols is m,
Figure FDA0002427175240000027
when k is<When m is greater than the total number of the carbon atoms,
Figure FDA0002427175240000028
wherein the content of the first and second substances,
Figure FDA0002427175240000029
is the input sum of the ith element of the kth layer;
Figure FDA00024271752400000210
is the output of the ith element of the kth layer;
Figure FDA00024271752400000211
the connection weight value from the ith element of the kth-1 layer to the jth element of the kth layer is obtained; f is an excitation function, which may be Sc(x)；
Figure FDA00024271752400000212
Is the actual output of the jth element of the mth layer relative to the weight vector W and the input vector X; y isiIs the desired output of the jth element of the mth layer associated with the weight vector W and the input vector X, wherein the mth layer is the output layer.
5. The method according to claim 1, wherein the pre-construction safety prediction neural network model and the in-construction safety prediction neural network model are constructed with a maximum number of iterations set to 5000 and a learning rate η set to 0.5.
6. The method of claim 1, wherein the pre-construction safety prediction neural network performance is determined to be stable when an error rate of the pre-construction safety prediction neural network model and the in-construction safety prediction neural network model is less than a preset value.
7. The multi-factor safety prediction system for the whole process of underground large space construction is characterized by comprising at least one processor and a memory which is in communication connection with the at least one processor; the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1 to 6.
CN202010224430.7A 2020-03-26 2020-03-26 Multi-factor safety stage prediction method for whole process of underground large space construction Active CN111582632B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010224430.7A CN111582632B (en) 2020-03-26 2020-03-26 Multi-factor safety stage prediction method for whole process of underground large space construction

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010224430.7A CN111582632B (en) 2020-03-26 2020-03-26 Multi-factor safety stage prediction method for whole process of underground large space construction

Publications (2)

Publication Number Publication Date
CN111582632A true CN111582632A (en) 2020-08-25
CN111582632B CN111582632B (en) 2023-03-28

Family

ID=72120549

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010224430.7A Active CN111582632B (en) 2020-03-26 2020-03-26 Multi-factor safety stage prediction method for whole process of underground large space construction

Country Status (1)

Country Link
CN (1) CN111582632B (en)

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1408385A1 (en) * 2002-10-11 2004-04-14 STMicroelectronics S.A. Method for controlling a dynamic system, using a fuzzy logic model of at least one inverse transfer function of the system
CN1752356A (en) * 2005-08-11 2006-03-29 西安理工大学 Intelligent model constructuring method for analyzing cavity wall rock stability
WO2010128956A2 (en) * 2009-05-08 2010-11-11 Inci Sengezer Innovation in building construction technology with pre-fabricated elements
CN106874581A (en) * 2016-12-30 2017-06-20 浙江大学 A kind of energy consumption of air conditioning system in buildings Forecasting Methodology based on BP neural network model
CN108535434A (en) * 2018-04-09 2018-09-14 重庆交通大学 Method based on Neural Network model predictive building site surrounding body turbidity
CN108920865A (en) * 2018-07-20 2018-11-30 广东工业大学 A kind of damage concrete structure restorative procedure, system and equipment and storage medium
CN109063403A (en) * 2018-10-22 2018-12-21 西安石油大学 A kind of slippery water Optimized fracturing design method
CN109636010A (en) * 2018-11-23 2019-04-16 国网湖北省电力有限公司 Provincial power network short-term load forecasting method and system based on correlative factor matrix
CN109816158A (en) * 2019-01-04 2019-05-28 平安科技（深圳）有限公司 Combined method, device, equipment and the readable storage medium storing program for executing of prediction model
CN110610226A (en) * 2018-06-14 2019-12-24 北京德知航创科技有限责任公司 Generator fault prediction method and device

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1408385A1 (en) * 2002-10-11 2004-04-14 STMicroelectronics S.A. Method for controlling a dynamic system, using a fuzzy logic model of at least one inverse transfer function of the system
CN1752356A (en) * 2005-08-11 2006-03-29 西安理工大学 Intelligent model constructuring method for analyzing cavity wall rock stability
WO2010128956A2 (en) * 2009-05-08 2010-11-11 Inci Sengezer Innovation in building construction technology with pre-fabricated elements
CN106874581A (en) * 2016-12-30 2017-06-20 浙江大学 A kind of energy consumption of air conditioning system in buildings Forecasting Methodology based on BP neural network model
CN108535434A (en) * 2018-04-09 2018-09-14 重庆交通大学 Method based on Neural Network model predictive building site surrounding body turbidity
CN110610226A (en) * 2018-06-14 2019-12-24 北京德知航创科技有限责任公司 Generator fault prediction method and device
CN108920865A (en) * 2018-07-20 2018-11-30 广东工业大学 A kind of damage concrete structure restorative procedure, system and equipment and storage medium
CN109063403A (en) * 2018-10-22 2018-12-21 西安石油大学 A kind of slippery water Optimized fracturing design method
CN109636010A (en) * 2018-11-23 2019-04-16 国网湖北省电力有限公司 Provincial power network short-term load forecasting method and system based on correlative factor matrix
CN109816158A (en) * 2019-01-04 2019-05-28 平安科技（深圳）有限公司 Combined method, device, equipment and the readable storage medium storing program for executing of prediction model

Non-Patent Citations (8)

* Cited by examiner, † Cited by third party
Title
侯跃强: "串联式组合模型在基坑监测数据分析及预测中的实践探究", 《资源信息与工程》 *
华博深: "灰色组合模型在基坑监测数据处理中的运用", 《中国优秀博硕士学位论文全文数据库(硕士)工程科技Ⅱ辑》 *
孙学聪: "深基坑变形监测及变形预测研究", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 *
张向君: "《信息分析与数据统计学习》", 28 February 2009 *
武鹏: "变形监测多模型组合预测方法研究", 《铁道勘察》 *
汪学清 等: "BP网络并串联模型用于施工质量管理评价的研究", 《建筑技术》 *
陈浩 等: "串联式组合模型在基坑监测中的应用", 《盐城工学院学报(自然科学版)》 *
雷升祥 等: "城市地下空间开发利用现状及未来发展理念", 《地下空间与工程学报》 *

Also Published As

Publication number Publication date
CN111582632B (en) 2023-03-28

Similar Documents

Publication Publication Date Title
Storlie et al. Implementation and evaluation of nonparametric regression procedures for sensitivity analysis of computationally demanding models
CN107563067A (en) Analysis of structural reliability method based on Adaptive proxy model
CN112926273A (en) Method for predicting residual life of multivariate degradation equipment
Zhou et al. Degradation evaluation of lateral story stiffness using HLA-based deep learning networks
CN111784061B (en) Training method, device and equipment for power grid engineering cost prediction model
WO2019160138A1 (en) Causality estimation device, causality estimation method, and program
CN115935834A (en) History fitting method based on deep autoregressive network and continuous learning strategy
CN114492211A (en) Residual oil distribution prediction method based on autoregressive network model
CN117076931B (en) Time sequence data prediction method and system based on conditional diffusion model
CN113537614A (en) Construction method, system, equipment and medium of power grid engineering cost prediction model
CN110852415B (en) Vegetation index prediction method, system and equipment based on neural network algorithm
CN111582632B (en) Multi-factor safety stage prediction method for whole process of underground large space construction
CN111582634A (en) Multi-factor safety grading method and system for underground large-space construction
Segura et al. Expected seismic performance of gravity dams using machine learning techniques
CN116502539A (en) VOCs gas concentration prediction method and system
Gaspar et al. Implicit material modelling using artificial intelligence techniques
Ribeiro Machado da Silva et al. Convolutional Neural Networks Applied to Flexible Pipes for Fatigue Calculations
CN112862211A (en) Method and device for assigning orders of dynamic ring defects of communication management system
CN113988210A (en) Method and device for restoring distorted data of structure monitoring sensor network and storage medium
CN112560252A (en) Prediction method for residual life of aircraft engine
Paiva et al. Quantum-inspired optimization of echo state networks applied to system identification
CN116700206B (en) Industrial control system anomaly detection method and device based on multi-modal neural network
CN116303786B (en) Block chain financial big data management system based on multidimensional data fusion algorithm
CN116362628B (en) Method and device for analyzing and evaluating running state of power system
Najar et al. Comparative Machine Learning Study for Estimating Peak Cladding Temperature in AP1000 Under LOFW

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