CN110442979B - BP neural network-based shield construction tunnel total deformation prediction method and system - Google Patents
BP neural network-based shield construction tunnel total deformation prediction method and system Download PDFInfo
- Publication number
- CN110442979B CN110442979B CN201910730664.6A CN201910730664A CN110442979B CN 110442979 B CN110442979 B CN 110442979B CN 201910730664 A CN201910730664 A CN 201910730664A CN 110442979 B CN110442979 B CN 110442979B
- Authority
- CN
- China
- Prior art keywords
- deformation
- deformation prediction
- weight
- neural network
- candidate
- 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
Images
Classifications
-
- E—FIXED CONSTRUCTIONS
- E21—EARTH DRILLING; MINING
- E21D—SHAFTS; TUNNELS; GALLERIES; LARGE UNDERGROUND CHAMBERS
- E21D11/00—Lining tunnels, galleries or other underground cavities, e.g. large underground chambers; Linings therefor; Making such linings in situ, e.g. by assembling
- E21D11/04—Lining with building materials
- E21D11/08—Lining with building materials with preformed concrete slabs
-
- E—FIXED CONSTRUCTIONS
- E21—EARTH DRILLING; MINING
- E21D—SHAFTS; TUNNELS; GALLERIES; LARGE UNDERGROUND CHAMBERS
- E21D9/00—Tunnels or galleries, with or without linings; Methods or apparatus for making thereof; Layout of tunnels or galleries
- E21D9/06—Making by using a driving shield, i.e. advanced by pushing means bearing against the already placed lining
-
- 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
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- 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
- G06N3/084—Backpropagation, e.g. using gradient descent
Abstract
The disclosure provides a method and a system for predicting full deformation of a shield construction tunnel based on a BP neural network. The total deformation prediction method comprises the steps of obtaining historical data corresponding to candidate total deformation prediction indexes and corresponding segment deformation; calculating subjective weight of the candidate total deformation prediction indexes by adopting an analytic hierarchy process, calculating objective weight of the candidate total deformation prediction indexes by adopting a rough set theory, and calculating combination weight of the candidate total deformation prediction indexes according to the difference degree of the subjective weight and the objective weight; screening candidate full-deformation prediction indexes corresponding to the combination weights larger than or equal to a preset threshold value to serve as full-deformation prediction indexes; multiplying historical data corresponding to all full-deformation prediction indexes by corresponding combination weights to obtain training sample data of the BP neural network, and training the BP neural network; and acquiring real-time data corresponding to the total deformation prediction index, multiplying the acquired real-time data by the corresponding combination weight, inputting the multiplied real-time data into the trained BP neural network, and outputting the segment deformation.
Description
Technical Field
The disclosure belongs to the field of prediction of tunnel total deformation, and particularly relates to a method and a system for predicting total deformation of a shield construction tunnel based on a BP neural network.
Background
The statements in this section merely provide background information related to the present disclosure and may not necessarily constitute prior art.
The underwater tunnel engineering has the advantages of stronger war damage resistance, small influence of natural disasters and the like; the method has the characteristics of no damage to shipping, no influence on ecological environment of sea areas and the like, and is widely applied to the project of crossing the sea and crossing the river. The shield construction method is the most advanced tunnel construction technology at present, and is always used as the main construction method of the underwater tunnel due to the advantages of high construction speed, high safety, small environmental disturbance and the like. However, the shield construction still has potential safety hazards due to the complex and special construction environment of the underwater tunnel.
The same as the traditional tunnel construction, the problem of overlarge deformation of a tunnel supporting structure is also faced in the shield method construction process. For the shield method construction tunnel, the segment structure is the main supporting structure of the tunnel. The displacement of the segments is too large, so that the dislocation among the segments can be caused, the connecting bolts are sheared or even sheared, and the structure is damaged; meanwhile, the damage of bolts for connecting the segments can be caused by the overlarge deformation of the segments, so that the damage of a structural waterproof layer is caused, and the quality problems of water leakage and the like of a tunnel are caused; in the construction process of the underwater tunnel, due to the combined action of self gravity and buoyancy on the pipe piece, the overlying soil can be locally compressed and cracked to form a through crack, so that some impermeable strata become permeable layers, and meanwhile, the overlying soil is subjected to the buoyancy effect to further deform the pipe piece of the tunnel. If the abnormal deformation problem of the duct piece cannot be found in time, duct piece cracks or even breakage can be caused when the additional bending moment generated by the eccentric force of the surrounding rock of the duct piece is serious. Therefore, proper measures are taken to control the displacement of the shield tunnel segment, so that the line type of the tunnel can be ensured to meet the design requirement, and the key point of ensuring the building clearance and the construction quality of the tunnel is also realized. In the process of tunnel excavation and supporting system construction, monitoring and measuring are important means for mastering the dynamic change process of surrounding rocks in tunnel construction, and the data obtained by monitoring and measuring can directly express the dynamic change and supporting condition in a certain period of time in a tunnel. And the bearing-deformation-time characteristics of the surrounding rock and the supporting structure can be determined by monitoring and measuring.
The deformation problem in the traditional tunnel construction process is inevitable, and the same is true for constructing the tunnel by a shield method. Due to the limitation of the construction section inside the shield tunneling machine, sensor elements and monitoring points for monitoring and measuring cannot be arranged in a certain distance behind the excavation face. The inventor finds that deformation information of a tunnel cannot be timely acquired in a plurality of sections in the tunnel, so that the deformation of a segment in a certain section is too large and cannot be timely monitored. Especially for the construction of an underwater tunnel shield method, the direct and indirect loss and the consequences caused by overlarge deformation of the pipe piece are immeasurable.
Disclosure of Invention
In order to solve the above problems, a first aspect of the present disclosure provides a full deformation prediction method for a shield construction tunnel based on a BP neural network, which considers the reasons of experience and cognition, calculates subjective weights of candidate full deformation prediction indexes by using an analytic hierarchy process, calculates objective weights of the candidate full deformation prediction indexes by using a rough set theory, calculates combined weights of the candidate full deformation prediction indexes according to the difference degree between the subjective weights and the objective weights, reduces the candidate full deformation prediction indexes, determines the full deformation prediction indexes to serve as input nodes of the BP neural network model, outputs segment deformation, and improves the accuracy of the full deformation prediction result of the shield construction tunnel.
In order to achieve the purpose, the following technical scheme is adopted in the disclosure:
a full deformation prediction method of a shield construction tunnel based on a BP neural network comprises the following steps:
acquiring historical data corresponding to the candidate total deformation prediction indexes and corresponding segment deformation;
calculating subjective weight of the candidate total deformation prediction indexes by adopting an analytic hierarchy process, calculating objective weight of the candidate total deformation prediction indexes by adopting a rough set theory, and calculating combination weight of the candidate total deformation prediction indexes according to the difference degree of the subjective weight and the objective weight;
screening candidate full-deformation prediction indexes corresponding to the combination weights larger than or equal to a preset threshold value to serve as full-deformation prediction indexes;
multiplying historical data corresponding to all full-deformation prediction indexes by corresponding combination weights to obtain training sample data of the BP neural network, and further constructing a training sample set and training the BP neural network;
and acquiring real-time data corresponding to the total deformation prediction index, multiplying the acquired real-time data by the corresponding combination weight, inputting the multiplied real-time data into the trained BP neural network, and outputting the segment deformation.
A second aspect of the present disclosure provides a full deformation prediction system for a shield construction tunnel based on a BP neural network.
A full deformation prediction system of a shield construction tunnel based on a BP neural network comprises:
the candidate full-deformation prediction index data acquisition module is used for acquiring historical data corresponding to the candidate full-deformation prediction index and corresponding segment deformation;
the combination weight calculation module is used for calculating the subjective weight of the candidate total-deformation prediction indexes by adopting an analytic hierarchy process, calculating the objective weight of the candidate total-deformation prediction indexes by adopting a rough set theory, and calculating the combination weight of the candidate total-deformation prediction indexes according to the difference degree of the subjective weight and the objective weight;
the total-deformation prediction index screening module is used for screening candidate total-deformation prediction indexes corresponding to the combination weights which are greater than or equal to a preset threshold value as total-deformation prediction indexes;
the BP neural network training module is used for multiplying historical data corresponding to all the full-deformation prediction indexes by corresponding combination weights to obtain training sample data of the BP neural network, and further constructing a training sample set and training the BP neural network;
and the segment deformation prediction module is used for acquiring real-time data corresponding to the total deformation prediction index, multiplying the acquired real-time data by the corresponding combination weight, inputting the multiplied real-time data into the trained BP neural network, and outputting segment deformation.
A third aspect of the present disclosure provides a computer-readable storage medium.
A computer readable storage medium, on which a computer program is stored, which when executed by a processor implements the steps in the method for predicting full deformation of a shield construction tunnel based on a BP neural network as described above.
A fourth aspect of the disclosure provides a computer terminal.
A computer terminal comprises a memory, a processor and a computer program which is stored on the memory and can run on the processor, wherein the processor executes the program to realize the steps of the full deformation prediction method of the shield construction tunnel based on the BP neural network.
The beneficial effects of this disclosure are:
in consideration of experience and cognition reasons, the method adopts an analytic hierarchy process to calculate subjective weight of a candidate total-deformation prediction index, adopts a rough set theory to calculate objective weight of the candidate total-deformation prediction index, calculates combination weight of the candidate total-deformation prediction index according to the difference degree of the subjective weight and the objective weight, reduces the candidate total-deformation prediction index, determines the total-deformation prediction index as an input node of a BP neural network model, outputs segment deformation, and improves the accuracy of a total-deformation prediction result of the shield construction tunnel.
Drawings
The accompanying drawings, which are included to provide a further understanding of the disclosure, illustrate embodiments of the disclosure and together with the description serve to explain the disclosure and are not to limit the disclosure.
Fig. 1 is a flowchart of a full deformation prediction method for a shield construction tunnel based on a BP neural network according to an embodiment of the present disclosure.
Fig. 2 is a schematic diagram of deformation parameters of a tube sheet provided by an embodiment of the present disclosure.
Fig. 3 is a temporal full-deformation curve provided by an embodiment of the present disclosure.
Fig. 4 is a schematic structural diagram of a total deformation prediction system of a shield construction tunnel based on a BP neural network according to an embodiment of the present disclosure.
Detailed Description
The present disclosure is further described with reference to the following drawings and examples.
It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the disclosure. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of example embodiments according to the present disclosure. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, and it should be understood that when the terms "comprises" and/or "comprising" are used in this specification, they specify the presence of stated features, steps, operations, devices, components, and/or combinations thereof, unless the context clearly indicates otherwise.
Example 1
The embodiment provides a full deformation prediction method for a shield construction tunnel based on a BP neural network, as shown in fig. 1, the method specifically includes:
s101: and acquiring historical data corresponding to the candidate total deformation prediction indexes and corresponding segment deformation.
Deformation monitoring during tunnel construction is an important basis for safety control and construction guidance in the construction process, monitoring points are distributed in a tunnel in the engineering construction process, a total station is adopted for monitoring and measuring, and original data are processed to obtain final tunnel deformation data. The content of engineering monitoring measurationis mainly the deformation monitoring of tunnel segment structure, including segment vertical displacement monitoring and horizontal displacement monitoring.
Similar to the traditional new Austrian method construction, as a main supporting form of the whole tunnel, the duct piece forms a relatively stable supporting system in the laying process, so that the stability of the tunnel in the construction process is ensured. In the shield construction process, the propulsion of the shield machine and the laying of the segments can generate inevitable disturbance to surrounding rocks and soil bodies. The duct piece is used as a main supporting structure of the tunnel, and can generate vertical deformation and horizontal deformation under the action of uneven disturbance of surrounding rocks or soil, as shown in fig. 2.
By the time attitude total deformation can know, there is certain disturbance scope in the shield construction process to the country rock influence, and in the shield construction process, tunnel supporting construction-section of jurisdiction structure in a section distance in excavation face rear still can receive the construction influence, but the section of jurisdiction deformation degree of different lag distances is different. As shown in fig. 3, the shield machine is constructed in the same time period of ζ1,ζ2,ζ3...ζnReferred to as the temporal full deformation of the time period.
In particular implementations, the candidate total deformation predictors include, but are not limited to, tunnel burial depth, coverage-span ratio, natural density, elastic modulus, poisson's ratio, tunneling speed, and hysteresis distance.
(1) Depth of tunnel burial
For an underwater tunnel, strong water and soil pressure is a main influence factor of the stability of the tunnel construction process and is also an important influence factor of segment deformation. The deeper the tunnel is buried, the greater the load of the overlying water and soil is, and the longer the stabilization time is required for the deformation caused by the construction disturbance. In the embodiment, the tunnel burial depth in the actual working condition, that is, the distance from the upper surface of the tunnel, which is in contact with surrounding rocks, to the sea level (the distance from the non-sea area section to the earth surface) is used as the candidate total deformation prediction index.
(2) Span ratio e
For full-section tunnel construction, the larger the section size is, the larger the influence of the construction process on the tunnel stability may be. The covering and crossing ratio is the ratio of the thickness of the surrounding rock layer on the tunnel to the diameter of the tunnel. The coverage-span ratio is increased, and the disturbance to the water around the tunnel is increased in the tunnel excavation process. In shield construction tunnels, the ratio of overburden thickness to tunnel diameter is referred to.
(3) Natural density ρ
The natural density of the rock mass determines the stability of the rock mass to a certain extent, and has a certain relation with the formation period, type and strength of the rock mass, thereby influencing the stability of the surrounding rock of the underwater tunnel to a certain extent.
(4) Modulus of elasticity E
The elastic modulus is an important index influencing the stability of rock and soil mass, and the size of the elastic modulus reflects the deformation characteristic of the rock mass. And dividing the elastic modulus corresponding to different deformation grades according to the existing elastic modulus dividing standard.
(5) Poisson ratio mu
The poisson ratio is a deformation state parameter which changes along with a stress state and a loading mode, and the horizontal distribution of stress is influenced to a great extent by the value of the poisson ratio. The poisson ratio is an important parameter of rock deformation characteristics and has a remarkable influence on rock deformation.
(6) Tunneling speed V
And (3) carrying out slurry balance shield construction, wherein the tunneling speed under the normal construction tunneling condition is generally 20-50 mm/min. The highest tunneling speed of the slurry shield is 92mm/min at present. The tunneling speed of the shield tunneling machine shows the quality of surrounding rocks in the tunnel construction process on one hand, and meanwhile, the surrounding rocks are disturbed, and the higher the tunneling speed is, the higher the disturbance degree of the surrounding rocks is relatively, so that the actual construction speed is quantized and graded. See table 1:
TABLE 1 Shield construction tunneling status
(7) Hysteresis distance R
In the shield construction process, the disturbance to surrounding rocks around is large, and after the primary lining pipe piece is laid and grouting is completed, the disturbance to the surrounding rocks and the deformation of the pipe piece are controlled to a certain degree. The influence of shield construction on segment deformation is weakened along with the increase of the lag distance, and the lag distance is obvious, so that the lag distance can be regarded as an evaluation and prediction index of segment deformation caused by the influence of construction. As shown, where R is the construction hysteresis distance, the segment deformation after this distance is considered as hysteresis deformation. Discretizing the lag distance R in the range of 0-150 m according to the practical situation of the engineering project.
In consideration of the fact that the actual construction process cannot completely and effectively monitor the full deformation data, the method of numerical simulation or model experiment is adopted for obtaining the full deformation data. Numerical simulation methods are preferably employed. An initial sample data set N is obtained. According to the process total deformation curve, in the tunnel construction process, the segment deformation can go through more than three deformation stages, regression analysis is carried out according to the temporal total deformation curve, fitting is carried out according to an initial sample data set N obtained through numerical simulation or model experiment, meanwhile, different phases of two curve equations are adjusted according to comparison of actually measured data, and a sample data set N' after data correction is obtained. As a data sample for a full deformation prediction study.
The quantitative grading method is characterized in that the quantitative grading is conducted on the construction deformation of the shield tunnel on the basis of the tunnel deformation research in the current stage, the quantitative grading is conducted on the vertical displacement of the segment within 0-60 mm respectively according to the deformation requirement of the segment in the shield construction, and the horizontal convergence of the segment is conducted on the quantitative grading of 0-40 mm.
TABLE 2 segment deformation level grading
And (3) dividing segment deformation into 5 grades, wherein a grade set S is equal to a full deformation grade of the segment of the underwater tunnel constructed by the shield method S1,S2,S3,S4,S5And (5) establishing a grade space and determining the quantitative interval of each evaluation index in each grade. And specifies S1(ii) no significant distortion; s2-deformation is small;S3{ with significant distortion }; s4{ denaturation is large }; s5-maximum distortion; the stability is described for the grades in table 2.
S102: and calculating the subjective weight of the candidate total-deformation prediction indexes by adopting an analytic hierarchy process, calculating the objective weight of the candidate total-deformation prediction indexes by adopting a rough set theory, and calculating the combined weight of the candidate total-deformation prediction indexes according to the difference degree of the subjective weight and the objective weight.
In the specific implementation, the evaluation indexes are compared in pairs by using an analytic hierarchy process, a decision maker uses the suggestion of an expert scholarer for reference and takes the scale of 1-9 as the standard of index importance evaluation, the indexes are subjected to overall importance ranking, and a judgment matrix G is constructedn×n(wherein I)i→jThe importance degree of I index compared with j index is Iij)：
And the influence degree of the evaluation index on the decision result is expressed quantitatively by taking a scale of 1-9 as the importance degree of the index. Such as Ii→jThe importance degree of the I index and the j index is Ii→jThe same I index is compared with the j index, and the importance degree is 1/Ii→j. Judging matrix G by adopting maximum eigenvalue methodn×nFinding the maximum eigenvalue λmaxAnd corresponding feature vectors alpha are calculated by formulas (1) to (3), and importance ranking of the evaluation indexes is obtained.
G·α＝λmaxα (1)
Finally, in order to avoid the generation of ordering contradiction in the importance comparison process due to excessive indexes in the decision making process, a consistency ratio index CR is introduced, and the following are defined:
CR＝CI/RI (2)
the consistency index CI and the average consistency index RI are calculated according to the formula (4).
CI＝(λmax-n)/(n-1) (3)
Wherein the value of the consistency index RI is as shown in table 3:
TABLE 3 RI values
Defining a decision matrix based on the consistency discrimination factor when CR<When 0.1, the judgment matrix is considered to have acceptable consistency, and the weight w of the deformation prediction indexiThe determination may be made from the eigenvector corresponding to the largest eigenvalue:
wherein w is more than or equal to 0i1 or less, and
αiis the ith element corresponding to the weight vector.
And is considered to be when CR<When the weight value w is 0.1, judging that the matrix G has acceptable consistency, and obtaining the weight value w calculated by the analytic hierarchy processi(w1,w2,w3......wn). On the contrary, the judgment matrix G is considered to have too large deviation degree from the consistency, and the element values in G need to be modified.
The objective calculation method adopts a rough set calculation method for calculation. And preferentially constructing a full-deformation prediction system according to an underwater tunnel shield method. And establishing a decision table, and calculating the objective weight of each evaluation index by adopting a rough set calculation method. And obtaining an objective weight value of the prediction index by adopting a rough set theory. The specific calculation steps are as follows:
let S ═ U, R be a knowledge expression system, where U is called a domain of discourse, representing a non-empty finite set of objects; r represents a condition attribute set. Wherein P, Q all belong to R, if R ═ Pgou Q is simultaneously present
Then C ═ (U, R, P, Q) is taken as the decision table.
P-to-Q support Sp(Q) can be calculated by the formula:
SP(Q)＝|posp(Q)|/|U| (0≤γP(Q)≤1) (5)
posp(Q) is represented as a normal region of P to Q. Gamma rayp(Q) indicates the degree of dependence of Q on P. Removing the index P from the condition attribute setiBesides, the other indexes have the support degree gamma to Qp-pi(Q) can be calculated by the formula:
for a subset P belonging to PiImportance to Q σPQ(Pi) Can be expressed as
PiWeight e ofiCan be expressed as:
obtaining a set of objective weights e from the coarse seti(e1,e2,e3......en)。
In order to ensure that the difference degree between the weight values obtained by the two methods is consistent with the difference degree of the corresponding distribution coefficients, a distance function method f (x) is introduced to express the difference degree of the weights of all indexes, and the distribution coefficients alpha and beta of the two weights are calculated by using formulas (9) to (13).
Let the combining weight be WiAnd the combined weight value is linearly weighted by the two:
Wi＝αwi+βei (10)
in the formula, alpha and beta are respectively distribution coefficients of two weights, in order to make the difference degree between different weights consistent with the difference degree between the distribution coefficients, the two formulas are made equal, and a constraint condition is added:
α+β＝1 (11)
in order to make the degree of difference between subjective and objective weights consistent with the degree of difference between distribution coefficients, the distance function should satisfy:
f(wi,ei)＝(α-β)2 (12)
in summary, the simultaneous equations set forth above, let:
the weights obtained by the two calculation methods are obtained through the joint type (9) to (11). Substituting the obtained distribution coefficient into equation set (13) to obtain final weight Wi. Analyzing the final weight to obtain the final total deformation prediction index and the weight W thereofi(W1,W2,W3......Wn). n represents the number of candidate full-deformation predictors.
S103: screening candidate full-deformation prediction indexes corresponding to the combination weights larger than or equal to a preset threshold value to serve as full-deformation prediction indexes;
for the weight calculation result, the weight is smaller than a preset threshold (for example, 0.1), if more weight exists, the calculation result with the minimum weight is reduced, and a reduced prediction index set G is obtainedi(G1,G2,G3......Gs) Wherein (s is less than or equal to n).
S104: and multiplying the historical data corresponding to each full-deformation prediction index by the corresponding combination weight to obtain training sample data of the BP neural network, and further constructing a training sample set and training the BP neural network.
In order to avoid the influence of unit dimensions of different prediction indexes on the prediction result, the input parameters are normalized according to the following formula.
Wherein g isijJ data, max (g), for the i indexi) Is the maximum value of the i-th index, min (g)i) Is the minimum value of the i-th index. bijAnd normalizing the result of the j data of the ith index.
And (5) attributing the parameters to the [0,1] interval, and taking the normalized data sample as input data.
Determining the structure of the neural network according to an empirical formula:
wherein h is the number of neurons in the hidden layer, i is the number of neurons in the input layer, and o is the number of neurons in the output layer.
An appropriate training function is selected for training. This position is the first layer transfer function, tan sig by default. The second layer transfer function, preferably, is thingdx. The momentum gradient is decreased to improve the training function lerngdm.
And setting the precision of model training, the learning error of the training process and the training step number according to the calculation requirement and the scale of the calculation data.
And after the training is finished, establishing a model for analyzing the calculation of the engineering example to obtain a prediction result y. And performing inverse normalization on the prediction result Y to obtain a final prediction result Y. And verifying and perfecting the prediction model by analyzing the prediction result.
Y＝y*(ymax-ymin)+ymin (17)
S105: and acquiring real-time data corresponding to the total deformation prediction index, multiplying the acquired real-time data by the corresponding combination weight, inputting the multiplied real-time data into the trained BP neural network, and outputting the segment deformation.
In another embodiment, the method further comprises:
and obtaining the deformation grade corresponding to the deformation of the duct piece according to the division of the preset duct piece deformation and the corresponding deformation grade.
Example 2
As shown in fig. 4, the system for predicting total deformation of a shield construction tunnel based on a BP neural network of the present embodiment includes:
(1) the candidate full-deformation prediction index data acquisition module is used for acquiring historical data corresponding to the candidate full-deformation prediction index and corresponding segment deformation;
in the candidate full-deformation prediction index data acquisition module, the candidate full-deformation prediction indexes include but are not limited to tunnel burial depth, coverage-span ratio, natural density, elastic modulus, poisson ratio, tunneling speed and lag distance.
(2) The combination weight calculation module is used for calculating the subjective weight of the candidate total-deformation prediction indexes by adopting an analytic hierarchy process, calculating the objective weight of the candidate total-deformation prediction indexes by adopting a rough set theory, and calculating the combination weight of the candidate total-deformation prediction indexes according to the difference degree of the subjective weight and the objective weight;
in the combination weight calculation module, the combination weight of the candidate total deformation prediction index is as follows:
Wi＝αwi+βei
wherein, WiIs the combined weight; f (w)i,ei) Representing distance functions, referring to subjective weights wiAnd objective weight eiThe degree of difference of (a); alpha, beta respectively represent subjective weight wiAnd objective weight eiThe distribution coefficient of (a); n represents the number of candidate full-deformation predictors.
(3) The total-deformation prediction index screening module is used for screening candidate total-deformation prediction indexes corresponding to the combination weights which are greater than or equal to a preset threshold value as total-deformation prediction indexes;
(4) the BP neural network training module is used for multiplying historical data corresponding to all the full-deformation prediction indexes by corresponding combination weights to obtain training sample data of the BP neural network, and further constructing a training sample set and training the BP neural network;
(5) and the segment deformation prediction module is used for acquiring real-time data corresponding to the total deformation prediction index, multiplying the acquired real-time data by the corresponding combination weight, inputting the multiplied real-time data into the trained BP neural network, and outputting segment deformation.
In another embodiment, the system further comprises:
and the deformation grade output module is used for obtaining the deformation grade corresponding to the segment deformation according to the division of the preset segment deformation and the corresponding deformation grade.
Example 3
The present embodiment provides a computer-readable storage medium, on which a computer program is stored, which when executed by a processor implements the steps in the full-deformation prediction method for a shield construction tunnel based on a BP neural network as shown in fig. 1.
Example 4
The embodiment provides a computer terminal, which comprises a memory, a processor and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the program to implement the steps in the method for predicting the total deformation of the shield construction tunnel based on the BP neural network shown in fig. 1.
As will be appreciated by one skilled in the art, embodiments of the present disclosure may be provided as a method, system, or computer program product. Accordingly, the present disclosure may take the form of a hardware embodiment, a software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present disclosure may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, optical storage, and the like) having computer-usable program code embodied therein.
The present disclosure is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the disclosure. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by a computer program, which can be stored in a computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), or the like.
The above description is only a preferred embodiment of the present disclosure and is not intended to limit the present disclosure, and various modifications and changes may be made to the present disclosure by those skilled in the art. Any modification, equivalent replacement, improvement and the like made within the spirit and principle of the present disclosure should be included in the protection scope of the present disclosure.
Claims (4)
1. A full deformation prediction method of a shield construction tunnel based on a BP neural network is characterized by comprising the following steps:
acquiring historical data corresponding to the candidate total deformation prediction indexes and corresponding segment deformation;
the candidate total deformation prediction indexes comprise tunnel burial depth, coverage-span ratio, natural density, elastic modulus, Poisson ratio, tunneling speed and lag distance;
calculating subjective weight of the candidate total deformation prediction indexes by adopting an analytic hierarchy process, calculating objective weight of the candidate total deformation prediction indexes by adopting a rough set theory, and calculating combination weight of the candidate total deformation prediction indexes according to the difference degree of the subjective weight and the objective weight;
the combination weight of the candidate total deformation prediction indexes is as follows:
Wi＝αwi+βei
wherein, WiIs the combined weight; f (w)i,ei) Representing distance functions, referring to subjective weights wiAnd objective weight eiThe degree of difference of (a); alpha, beta respectively represent subjective weight wiAnd objective weight eiThe distribution coefficient of (a); n represents the number of candidate total deformation prediction indexes;
screening candidate full-deformation prediction indexes corresponding to the combination weights larger than or equal to a preset threshold value to serve as full-deformation prediction indexes;
multiplying historical data corresponding to all full-deformation prediction indexes by corresponding combination weights to obtain training sample data of the BP neural network, and further constructing a training sample set and training the BP neural network;
acquiring real-time data corresponding to the total deformation prediction index, multiplying the acquired real-time data by the corresponding combination weight, inputting the multiplied real-time data into a trained BP neural network, and outputting the segment deformation;
and obtaining the deformation grade corresponding to the deformation of the duct piece according to the division of the preset duct piece deformation and the corresponding deformation grade.
2. A full deformation prediction system of a shield construction tunnel based on a BP neural network is characterized by comprising the following components:
the candidate full-deformation prediction index data acquisition module is used for acquiring historical data corresponding to the candidate full-deformation prediction index and corresponding segment deformation;
in the candidate full-deformation prediction index data acquisition module, the candidate full-deformation prediction indexes comprise tunnel burial depth, coverage-span ratio, natural density, elastic modulus, Poisson ratio, tunneling speed and lag distance;
the combination weight calculation module is used for calculating the subjective weight of the candidate total-deformation prediction indexes by adopting an analytic hierarchy process, calculating the objective weight of the candidate total-deformation prediction indexes by adopting a rough set theory, and calculating the combination weight of the candidate total-deformation prediction indexes according to the difference degree of the subjective weight and the objective weight;
in the combination weight calculation module, the combination weight of the candidate total deformation prediction index is as follows:
Wi＝αwi+βei
wherein, WiIs the combined weight; f (w)i,ei) Representing distance functions, referring to subjective weights wiAnd objective weight eiThe degree of difference of (a); alpha, beta respectively represent subjective weight wiAnd objective weight eiDistribution coefficient of(ii) a n represents the number of candidate total deformation prediction indexes;
the total-deformation prediction index screening module is used for screening candidate total-deformation prediction indexes corresponding to the combination weights which are greater than or equal to a preset threshold value as total-deformation prediction indexes;
the BP neural network training module is used for multiplying historical data corresponding to all the full-deformation prediction indexes by corresponding combination weights to obtain training sample data of the BP neural network, and further constructing a training sample set and training the BP neural network;
the segment deformation prediction module is used for acquiring real-time data corresponding to the total deformation prediction index, multiplying the acquired real-time data by the corresponding combination weight, inputting the multiplied real-time data into a trained BP neural network, and outputting segment deformation;
and the deformation grade output module is used for obtaining the deformation grade corresponding to the segment deformation according to the division of the preset segment deformation and the corresponding deformation grade.
3. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method for full-deformation prediction of a shield construction tunnel based on a BP neural network as set forth in claim 1.
4. A computer terminal comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor when executing the program implements the steps of the method for predicting full deformation of a shield construction tunnel based on a BP neural network as set forth in claim 1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910730664.6A CN110442979B (en) | 2019-08-08 | 2019-08-08 | BP neural network-based shield construction tunnel total deformation prediction method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910730664.6A CN110442979B (en) | 2019-08-08 | 2019-08-08 | BP neural network-based shield construction tunnel total deformation prediction method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110442979A CN110442979A (en) | 2019-11-12 |
| CN110442979B true CN110442979B (en) | 2021-04-13 |
Family
ID=68434031
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910730664.6A Active CN110442979B (en) | 2019-08-08 | 2019-08-08 | BP neural network-based shield construction tunnel total deformation prediction method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110442979B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112146748B (en) * | 2020-09-03 | 2022-12-23 | 深圳大学 | Method for positioning vibration source around tunnel by combining tunnel and earth surface monitoring data |
| CN112762853B (en) * | 2020-12-28 | 2021-11-16 | 中国科学院武汉岩土力学研究所 | Method and equipment for monitoring full deformation of duct piece in tunnel shield process |
| CN113240095B (en) * | 2021-06-07 | 2022-04-22 | 北京理工大学 | Casting cylinder cover mechanical property prediction method based on rough set and neural network |
| CN115217152B (en) * | 2022-07-29 | 2023-05-16 | 招商局重庆交通科研设计院有限公司 | Method and device for predicting open-close deformation of immersed tunnel pipe joint |
| CN117235874B (en) * | 2023-11-14 | 2024-02-20 | 中电建铁路建设投资集团有限公司 | Track deformation prediction method and system based on shield tunneling |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103093400A (en) * | 2013-01-24 | 2013-05-08 | 华中科技大学 | Adjacent building safety quantitative evaluation method in tunnel construction |
| CN104766129A (en) * | 2014-12-31 | 2015-07-08 | 华中科技大学 | Subway shield construction surface deformation warning method based on temporal and spatial information fusion |
| EP3080395A2 (en) * | 2013-12-13 | 2016-10-19 | SWS Engineering S.p.A. | Procedure for the construction of cross passages in double pipe tunnels |
| CN106547986A (en) * | 2016-11-08 | 2017-03-29 | 苏州大学 | A kind of tunnel soil pressure load computational methods |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| NL1015324C2 (en) * | 1999-11-11 | 2001-05-14 | Ballast Nedam Infra B V | Device and method for drilling in a substrate. |
| GB0316002D0 (en) * | 2003-07-09 | 2003-08-13 | Univ Aston | Sensing system and method |
| US7508701B1 (en) * | 2006-11-29 | 2009-03-24 | The Board Of Trustees Of The Leland Stanford Junior University | Negative differential resistance devices and approaches therefor |
| CN101344389B (en) * | 2008-08-20 | 2010-12-01 | 中国建筑第八工程局有限公司 | Method for estimating tunnel surrounding rock displacement by neural network |
| CN107092990A (en) * | 2017-05-03 | 2017-08-25 | 西安电子科技大学 | The shield construction ground settlement forecast system and method analyzed based on big data |
-
2019
- 2019-08-08 CN CN201910730664.6A patent/CN110442979B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103093400A (en) * | 2013-01-24 | 2013-05-08 | 华中科技大学 | Adjacent building safety quantitative evaluation method in tunnel construction |
| EP3080395A2 (en) * | 2013-12-13 | 2016-10-19 | SWS Engineering S.p.A. | Procedure for the construction of cross passages in double pipe tunnels |
| CN104766129A (en) * | 2014-12-31 | 2015-07-08 | 华中科技大学 | Subway shield construction surface deformation warning method based on temporal and spatial information fusion |
| CN106547986A (en) * | 2016-11-08 | 2017-03-29 | 苏州大学 | A kind of tunnel soil pressure load computational methods |
Non-Patent Citations (5)
| Title |
|---|
| A prediction model for overlying rock thickness of subsea tunnel: A hybrid intelligent system;Yiguo Xue et al;《Marine Georesources & Geotechnology》;20190208;第1-10页 * |
| Longitudinaljetventilationcalculationandapplicationoflonghighwaytunnel;BinghuaZhou et al;《JOURNAL OF ENGINEERING DESIGN AND TECHNOLOGY 》;20181231;第16卷(第2期);第256-269页 * |
| 南昌上软下硬地层中盾构施工地表沉降的BP神经网络预测方法;周纯择 等;《防灾减灾工程学报》;20150831;第35卷(第4期);第556-562页 * |
| 基于PSR－改进模糊集对分析模型的河流健康评价;周振民 等;《中国农村水利水电》;20181231(第12期);第77-81、86页 * |
| 盾构下穿施工对既有立交桥桩基的影响分析;吴瑞;《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》;20180415;第C034-554页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110442979A (en) | 2019-11-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110442979B (en) | BP neural network-based shield construction tunnel total deformation prediction method and system | |
| CN112879025B (en) | Intelligent regulation and control method for earth pressure balance shield earth pressure based on GRU deep learning | |
| CN112733221A (en) | Optimization design method and system for reinforcing side slope by using anti-slide pile | |
| CN110705795A (en) | Tunnel gush risk grade fuzzy prediction method based on comprehensive advanced geological prediction technology | |
| CN111157346B (en) | Analysis system and method for design and correction of water curtain system of underground water-sealed oil storage | |
| CN116629712B (en) | Submarine shield tunnel construction quality risk assessment method based on PSO-BP neural network | |
| CN110378574A (en) | Submerged tunnel Pressure Shield Tunnel face stability evaluation method, system and equipment | |
| CN115828374A (en) | Method and system for predicting non-uniform deformation of composite stratum tunnel in construction period | |
| CN115423167A (en) | Subway deep foundation pit construction safety early warning and decision-making assisting method and system | |
| CN114372393A (en) | Arch dam safety early warning method under influence of swath deformation | |
| CN106568647A (en) | Nerve network-based concrete strength predication method | |
| CN111666624A (en) | Method for evaluating stability of surrounding rock of large-span underground cavern under non-support condition | |
| CN115659749A (en) | Foundation pit deformation prediction method and system, electronic equipment and storage medium | |
| Hassan et al. | Predictive models to evaluate the interaction effect of soil-tunnel interaction parameters on surface and subsurface settlement | |
| CN113326660B (en) | Tunnel surrounding rock extrusion deformation prediction method based on GA-XGboost model | |
| CN116861704B (en) | Dynamic rapid prediction method for large deformation grade of high-ground-stress soft rock tunnel | |
| CN110348098B (en) | Clay loess tunnel excavation simulation parameter assignment method | |
| CN108446413B (en) | Method for optimally measuring pile diameter of grouting-formed club-footed pile | |
| CN113515840B (en) | Method for predicting rock mass excavation disturbance area and related equipment | |
| CN113221399B (en) | Decision method for dividing scheme of railway tunnel fabricated lining components | |
| CN115563874A (en) | Tunnel health degree analysis method and system based on big data | |
| CN106812031B (en) | A kind of heavy haul railway tunnel bottom construction design method considering train long term | |
| CN111929424B (en) | Large-span underground cavern hard surrounding rock sub-classification method based on size effect | |
| CN107577868A (en) | A kind of Optimization Design of over mountain line tunnel preliminary bracing | |
| CN113505428A (en) | Comprehensive quantitative design method for circular water-passing tunnel lining structure |
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 |