CN108416458A - A kind of tunnel rich water rock mass Synthetic Geological Prediction Ahead of Construction method based on BP neural network - Google Patents
A kind of tunnel rich water rock mass Synthetic Geological Prediction Ahead of Construction method based on BP neural network Download PDFInfo
- Publication number
- CN108416458A CN108416458A CN201810030866.5A CN201810030866A CN108416458A CN 108416458 A CN108416458 A CN 108416458A CN 201810030866 A CN201810030866 A CN 201810030866A CN 108416458 A CN108416458 A CN 108416458A
- Authority
- CN
- China
- Prior art keywords
- rich water
- rock mass
- assignment
- neural network
- water rock
- 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.)
- Pending
Links
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
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01V—GEOPHYSICS; GRAVITATIONAL MEASUREMENTS; DETECTING MASSES OR OBJECTS; TAGS
- G01V9/00—Prospecting or detecting by methods not provided for in groups G01V1/00 - G01V8/00
- G01V9/02—Determining existence or flow of underground water
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Software Systems (AREA)
- Development Economics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Computational Linguistics (AREA)
- Hydrology & Water Resources (AREA)
- General Life Sciences & Earth Sciences (AREA)
- Geophysics (AREA)
- Biophysics (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Game Theory and Decision Science (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Geophysics And Detection Of Objects (AREA)
Abstract
The tunnel rich water rock mass Synthetic Geological Prediction Ahead of Construction method based on BP neural network that the invention discloses a kind of, determines the master control influence factor of rich water rock mass;The division of risk class is carried out to rich water rock mass with geological researching data in conjunction with physical prospecting and drafts parameter area corresponding with each grade;Rich water rock mass prediction model and rich water rock mass impact factor and rich water grade quantification of the construction based on BP neural network;Neural network learning sample is built, and is learnt, the prediction model of neural network is obtained, tunnel rich water rock mass Synthetic Geological Prediction Ahead of Construction is carried out using prediction model.The present invention can be strong the site operation for instructing underground engineering, ensure the construction safety of one line of engineering.
Description
Technical field
The tunnel rich water rock mass Synthetic Geological Prediction Ahead of Construction method based on BP neural network that the present invention relates to a kind of.
Background technology
It with the continuous development of economic construction of China, needs to build a large amount of underground engineering, and was built in underground engineering
Cheng Zhong, rich water rock mass are most common a kind of unfavorable geologies during constructing tunnel, and a system can be caused to underground engineering construction
Row harm.Therefore, in order to ensure tunnel construction safety, the efficiency of engineering construction is improved, needs to have rich water rock mass
Effect exceeds the time limit to forecast.
Many methods are proposed for the forecast of rich water rock mass at present, mainly have geological analysis method and geophysical method two big
Class method.In order to improve the precision of prediction of rich water rock mass, the synthesis advance geologic that numerous scholars propose the prediction of rich water rock mass is pre-
Reporting method carries out the comprehensive forecasting of rich water rock mass using a variety of forecast means, achieves certain effect.But use this
When kind method carries out comprehensive forecasting, typically professional technician carries out manually comprehensive to the result that a variety of forecasting procedures are forecast
It closes analysis, judge.I.e. professional technician carries out artificial comprehensive to the forecast result of the method for a variety of forecast, relies on itself
Experience is summarized to a variety of forecast analysis results, is analyzed, obtains final forecast result.The accuracy of prediction result completely with
By means of the experience of expert, if the level of expert is preferably, that is, what is judged is accurate, if the level of expert is general, that is, judges to deposit
In higher mistake, the Comprehensive Evaluation of this rich water rock mass forecast, artificial subjectivity is strong, and it is objective to lack on Comprehensive Evaluation
Property.
BP (back propagation) neural network is 1986 by the science headed by Rumelhart and McClelland
The concept that family proposes is that a kind of reaction type connects multilayer neural network entirely, has stronger associative memory and Generalization Ability.Pass through
The mapping relations between outputting and inputting are established in study to learning sample, realize multiple parameters by input to predict it
Corresponding classification results.Standard BP model is made of (input layer, hidden layer and output layer) three layers of neuron.BP network learning procedures
It is divided into two stages, the first stage is signal forward-propagating process:After being supplied to network to mode of learning, the activation of neuron
It is worth from input layer through each hidden layer to output Es-region propagations, the input response of network is obtained in each neuron of output layer;Second-order
Section is error correction back-propagation process:If not obtaining desired output valve in output layer, calculate to step-by-step recursion practical defeated
Go out the error between desired output, by the direction for the error for reducing desired output and reality output, from output layer through each implicit
Layer each connection weight of layer-by-layer correction, eventually passes back to input layer.
Using a variety of method for forecasting advanced geology, the detection data of a variety of detection means of front of tunnel heading can get.BP god
The forecast result of a variety of method for forecasting advanced geology can be analyzed, is established between forecast result and unfavorable geology through network
Relationship.How Synthetic Geological Prediction Ahead of Construction achievement and BP neural network to be combined in Tunnel Engineering, to Tunnel Engineering area
Square rich water rock mass carries out advanced prediction in front, realizes the objective prediction to rich water rock mass, is a current skill urgently to be resolved hurrily
Art problem.
Invention content
The present invention is to solve the above-mentioned problems, it is proposed that a kind of tunnel rich water rock mass synthesis based on BP neural network is advanced
Geologic prediction method, the present invention is using the result of detection of Synthetic Geological Prediction Ahead of Construction as initial data, in conjunction with BP neural network technology
Tunnel tunnel face front rich water rock mass is accurately predicted, the disasters such as the Tunnel Landslide caused by rich water rock mass is reduced and occurs
Probability, effectively instruct the site operation of Tunnel Engineering, ensure the construction safety of one line of engineering.
To achieve the goals above, the present invention adopts the following technical scheme that:
A kind of tunnel rich water rock mass Synthetic Geological Prediction Ahead of Construction method based on BP neural network, includes the following steps:
(1) the master control influence factor of rich water rock mass is determined;
(2) it combines physical prospecting to carry out the division of risk class to rich water rock mass with geological researching data and drafts and each grade phase
Corresponding parameter area;
(3) rich water rock mass prediction model and rich water rock mass impact factor and rich water grade of the construction based on BP neural network
Quantification;
(4) neural network learning sample is built, and is learnt, the prediction model of neural network is obtained, utilizes prediction mould
Type carries out tunnel rich water rock mass Synthetic Geological Prediction Ahead of Construction.
In the step (1), the distinguishing indexes of rich water rock mass are determined:Design phase rich water information, face water outlet status,
Vp/Vs, electromagnetic waveform feature, change in apparent resistivity and/or resistivity.
In the step (2), according to the distinguishing indexes value of determining rich water rock mass, country rock rich water situation is divided into 4 kinds
Different types of output valve:Rich in underground water, contain medium underground water, anhydrous containing a small amount of underground water and drying.
In the step (3), the model structure of neural network is 3 layers, including input layer, middle layer and output layer, input
Layer inputted be rich water rock mass 6 distinguishing indexes, i.e. design phase rich water information, face water outlet status, Vp/Vs, electricity
Magnetic wave wave character, change in apparent resistivity and resistivity;Output layer output is corresponding desired output, i.e. the rich water feelings of rock mass
Condition.
In the step (3), the interstitial content of input layer, hidden layer and output layer is determined, using BP neural network algorithm
Establish rich water rock mass prediction model.
In the step (3), the output of neural network model is rich water situation, wherein being rich in underground water assignment 1;Containing a large amount of
Underground water assignment 2;Containing a small amount of underground water assignment 3;Dry anhydrous assignment 0.
Specifically, geological information is as the first input factor, the wherein very big assignment 1 of rich water possibility, rich water possibility compared with
Big assignment 2, possible aqueous assignment 3, substantially anhydrous assignment 0;
Face water outlet status is as the second input factor, wherein water drenching and strand water assignment 1, linear water assignment 2, humidity
Assignment of dripping 3, dry anhydrous assignment 0;
Vp/Vs inputs the factor as third, wherein very big assignment 1, larger assignment 2, bigger assignment 3, substantially anhydrous change
Change assignment 0;
Electromagnetic waveform feature is assigned as the 4th input factor, wherein discretization strong reflection amplitude wave assignment 1, noisy mode
Value 2, amplitude change is little, close to continuous lineups assignment 3, waveform short arc, high frequency, lineups assignment smoothly and continuously
4;
Change in apparent resistivity reduces larger assignment 2 as the 5th input factor wherein seriously reducing assignment 1, in a slight decrease
Assignment 3, steady unchanged assignment 0;
Resistivity situation is as the 6th input factor, wherein 0~0.035k of resistivity Ω m assignment 1, resistivity 0.035
~0.1k Ω m assignment 2, resistivity 0.1~0.275k Ω m assignment 3, resistivity are more than 0.275k Ω m assignment 4.
In the step (4), the typical case of collection research region rich water rock mass forecast builds learning sample, right later
Each connection weight assigns the random number in a section (0,1) respectively in neural network model, sets error function E, gives and calculates
Accuracy value and maximum study number M, network output is calculated according to input data.
In the step (4), corrected using the error term of each neuron of hidden layer and the input of each neuron of input layer
Hidden layer weights;The power of output unit is corrected using the error term of each neuron of output layer and the output of hidden layer each neuron
Value.
In the step (4), when being finished all samples, judge whether network error meets the requirements, when error reaches pre-
If precision or study number are more than default maximum times, then terminate algorithm;Otherwise, learn into next circle, until result reaches
Until default precision or study number are more than preset times.
In the step (4), the neural network prediction model of rich water rock mass is established according to the weights and threshold value obtained.
In the step (4), by the richness of the water outlet status and above-mentioned acquisition of the rich water situation of design phase and face
The geophysical prospecting informations of water rock mass are input in the rich water rock mass prediction model based on BP neural network, the calculating through neural network point
Analysis, judges the country rock rich water situation of front of tunnel heading.
Compared with prior art, beneficial effects of the present invention are:
The present invention is to be based on BP neural network technology, a kind of tunnel face that comprehensive a variety of advance geologic prediction data propose
The method of square rich water rock mass advanced prediction in front.This method is using the result of detection of a variety of method for forecasting advanced geology as original number
According to, tunnel tunnel face front rich water rock mass is accurately predicted in conjunction with BP neural network technology, improve to rich water rock mass
The objectivity and accuracy of prediction, the strong site operation for instructing underground engineering ensure the construction safety of one line of engineering.
Description of the drawings
The accompanying drawings which form a part of this application are used for providing further understanding of the present application, and the application's shows
Meaning property embodiment and its explanation do not constitute the improper restriction to the application for explaining the application.
Fig. 1 is the method for the present invention implementation flow chart.
Specific implementation mode:
The invention will be further described with embodiment below in conjunction with the accompanying drawings.
It is noted that following detailed description is all illustrative, it is intended to provide further instruction to the application.Unless another
It indicates, all technical and scientific terms used herein has usual with the application person of an ordinary skill in the technical field
The identical meanings of understanding.
It should be noted that term used herein above is merely to describe specific implementation mode, and be not intended to restricted root
According to the illustrative embodiments of the application.As used herein, unless the context clearly indicates otherwise, otherwise singulative
It is also intended to include plural form, additionally, it should be understood that, when in the present specification using term "comprising" and/or " packet
Include " when, indicate existing characteristics, step, operation, device, component and/or combination thereof.
In the present invention, term for example "upper", "lower", "left", "right", "front", "rear", "vertical", "horizontal", " side ",
The orientation or positional relationship of the instructions such as "bottom" is to be based on the orientation or positional relationship shown in the drawings, only to facilitate describing this hair
Bright each component or component structure relationship and the relative of determination, not refer in particular to either component or element in the present invention, cannot understand
For limitation of the present invention.
In the present invention, term such as " affixed ", " connected ", " connection " shall be understood in a broad sense, and indicate may be a fixed connection,
Can also be to be integrally connected or be detachably connected;It can be directly connected, it can also be indirectly connected through an intermediary.For
The related scientific research of this field or technical staff can determine the concrete meaning of above-mentioned term in the present invention as the case may be,
It is not considered as limiting the invention.
As shown in Figure 1, a kind of tunnel rich water rock mass method for forecasting advanced geology based on BP neural network technology, including with
Lower step:
Step 1:Determine the master control influence factor of rich water rock mass;
Various method for forecasting advanced geology are analyzed, the related geophysical prospecting method that can effectively identify rich water rock mass is found out.
Rich water rock mass Synthetic Geological Prediction Ahead of Construction method includes:Geological analysis method of prediction, TSP methods of prediction, geological radar method, BEAM methods.
By Literature Consult, all kinds of method for forecasting advanced geology of Integrated comparative can obtain with the relevant ginseng of rich water rock mass
Number, determines the distinguishing indexes of rich water rock mass:Design phase rich water information, face water outlet status, Vp/Vs, electromagnetic waveform are special
Sign, change in apparent resistivity, resistivity.
Step 2:The division of risk class is carried out to rich water rock mass in conjunction with physical prospecting and geological researching data, and draft with respectively
The corresponding parameter area of grade;
Country rock rich water situation is divided into 4 kinds of different types of output valves:Rich in underground water, contain medium underground water, containing few
Amount underground water, drying are anhydrous, and it is respectively design phase rich water information, area to choose 6 kinds of factors as the input value of prediction index
Face water outlet status, Vp/Vs, electromagnetic waveform feature, change in apparent resistivity, resistivity.
When rich water situation predicts that target is rich in underground water, the information of the six class factors is:Design phase rich water possibility pole
Greatly, face water drenching and strand water, Vp/Vs be very big, Vp stablizes unchanged, electromagnetic waveform discretization, strong reflection amplitude wave,
Apparent resistivity seriously reduces, resistivity 0~0.035k Ω m;
When rich water situation predicts that target is containing medium underground water, the information of the six class factors is:Design phase rich water possibility
Larger, the linear water of face, Vp/Vs be larger, electromagnetic waveform be clutter, apparent resistivity reduce larger, resistivity 0.035~
0.1kΩ·m;
When rich water situation predicts target as containing few equal underground water, the information of the six class factors is:Design phase may the aqueous, palm
Sub- upper thread shape water, Vp/Vs are bigger, electromagnetic waveform is that amplitude change is little, are slightly dropped close to continuous lineups, apparent resistivity
Low, resistivity 0.1~0.275k Ω m;
When rich water situation predicts target as containing few equal underground water, the information of the six class factors is:Design phase is substantially anhydrous, slaps
The drying of sub- face is anhydrous, Vp/Vs is substantially unchanged, electromagnetic waveform short arc, and high frequency is intensive, and lineups smoothly and continuously regard
The steady unchanged, resistivity of resistivity is more than 0.275k Ω m.
Step 3;Rich water rock mass prediction model and rich water rock mass impact factor and rich water of the construction based on BP neural network
Grade quantification;
The model structure of neural network is generally 3 layers, including input layer, middle layer and output layer.What input layer was inputted
For 6 distinguishing indexes of rich water rock mass, i.e. design phase rich water information, face water outlet status, Vp/Vs, electromagnetic waveform spy
Sign, change in apparent resistivity, resistivity.Output layer output is corresponding desired output, i.e. the rich water situation of rock mass.Middle layer section
Points are determined by empirical equation:
Wherein, each alphabetical meaning is:
m:Node in hidden layer, n:Input layer number, l:Output layer number of nodes, a:Constant between 1-10.
After the interstitial content for determining input layer, hidden layer and output layer, rich water rock mass is established using BP neural network algorithm
Prediction model.
Each impact factor and rich water grade to rich water rock mass carry out assignment, are convenient for the study and prediction of neural network.God
Output through network model is rich water situation, wherein being rich in underground water assignment 1;Containing a large amount of underground water assignment 2;Containing a small amount of underground water
Assignment 3;Dry anhydrous assignment 0.
The input of neural network model is 6 impact factors, and assignment situation is as follows:The geological information of design phase is as defeated
Enter the factor 1, the wherein very big assignment 1 of rich water possibility, the larger assignment 2 of rich water possibility, possible aqueous assignment 3, substantially anhydrous tax
Value 0;Face water outlet status is dripped tax as the input factor 2, wherein water drenching and strand water assignment 1, linear water assignment 2, humidity
Value 3, dry anhydrous assignment 0;Vp/Vs is used as the input factor 3, wherein very big assignment 1, larger assignment 2, bigger assignment 3, substantially
Anhydrous variation assignment 0.
Electromagnetic waveform feature is as the input factor 4, wherein discretization strong reflection amplitude wave assignment 1, noisy mode assignment
2, amplitude change is little, close to continuous lineups assignment 3, waveform short arc, high frequency, lineups assignment 4 smoothly and continuously;
Change in apparent resistivity reduces larger assignment 2 as the input factor 5 wherein seriously reducing assignment 1, assignment 3 in a slight decrease, steadily
Unchanged assignment 0;Resistivity situation is as the input factor 6, wherein 0~0.035k of resistivity Ω m assignment 1, resistivity 0.035
~0.1k Ω m assignment 2, resistivity 0.1~0.275k Ω m assignment 3, resistivity are more than 0.275k Ω m assignment 4.
Step 4:Neural network learning sample is built, and is learnt, the prediction model of neural network is obtained;
Its detailed process is:The typical case of collection research region rich water rock mass forecast first, builds learning sample, later
Random number in one section (0,1) is assigned respectively to each connection weight in neural network model, sets error function E, to devise a stratagem
Calculate accuracy value and maximum study number M.Network output is calculated according to input data afterwards.Utilize the error term of each neuron of hidden layer
Hidden layer weights are corrected in input with each neuron of input layer.Error term using each neuron of output layer and each god of hidden layer
The weights of output unit are corrected through the output of member.When being finished all samples, judge whether network error meets the requirements.When accidentally
Difference reaches default precision or study number is more than default maximum times, then terminates algorithm.Otherwise, learn into next circle.Until
As a result reach default precision or study number more than until preset times.Rich water rock mass is established according to the weights and threshold value obtained
Neural network prediction model.
Step 5:Engineer application.
To engineering site carry out geological analysis method advanced prediction, TSP methods advanced prediction, geology method radar advanced prediction,
BEAM method advanced predictions.
The data of detection are handled using corresponding the poster processing soft by professional technician, obtain Vp/Vs, electromagnetism
The information such as wave wave character, change in apparent resistivity, resistivity.
The physical prospecting of the rich water rock mass of the water outlet status and above-mentioned acquisition of the rich water situation of design phase and face is believed
Breath is input in the rich water rock mass prediction model based on BP neural network, the calculating analysis through neural network, before judging face
The country rock rich water situation of side.
The foregoing is merely the preferred embodiments of the application, are not intended to limit this application, for the skill of this field
For art personnel, the application can have various modifications and variations.Within the spirit and principles of this application, any made by repair
Change, equivalent replacement, improvement etc., should be included within the protection domain of the application.
Above-mentioned, although the foregoing specific embodiments of the present invention is described with reference to the accompanying drawings, not protects model to the present invention
The limitation enclosed, those skilled in the art should understand that, based on the technical solutions of the present invention, those skilled in the art are not
Need to make the creative labor the various modifications or changes that can be made still within protection scope of the present invention.
Claims (10)
1. a kind of tunnel rich water rock mass Synthetic Geological Prediction Ahead of Construction method based on BP neural network, it is characterized in that:Including following
Step:
(1) the master control influence factor of rich water rock mass is determined;
(2) it combines physical prospecting to carry out the division of risk class to rich water rock mass with geological researching data and drafts corresponding with each grade
Parameter area;
(3) rich water rock mass prediction model and rich water rock mass impact factor and rich water grade of the construction based on BP neural network are quantitative
Change;
(4) build neural network learning sample, and learnt, obtain the prediction model of neural network, using prediction model into
Row tunnel rich water rock mass Synthetic Geological Prediction Ahead of Construction.
2. a kind of tunnel rich water rock mass Synthetic Geological Prediction Ahead of Construction method based on BP neural network as described in claim 1,
It is characterized in that:In the step (1), the distinguishing indexes of rich water rock mass are determined:Design phase rich water information, face go out regimen
Condition, Vp/Vs, electromagnetic waveform feature, change in apparent resistivity and/or resistivity.
3. a kind of tunnel rich water rock mass Synthetic Geological Prediction Ahead of Construction method based on BP neural network as described in claim 1,
It is characterized in that:In the step (2), according to the distinguishing indexes value of determining rich water rock mass, country rock rich water situation is divided into 4 kinds
Different types of output valve:Rich in underground water, contain medium underground water, anhydrous containing a small amount of underground water and drying.
4. a kind of tunnel rich water rock mass Synthetic Geological Prediction Ahead of Construction method based on BP neural network as described in claim 1,
It is characterized in that:In the step (3), the model structure of neural network is 3 layers, including input layer, middle layer and output layer, input
Layer inputted be rich water rock mass 6 distinguishing indexes, i.e. design phase rich water information, face water outlet status, Vp/Vs, electricity
Magnetic wave wave character, change in apparent resistivity and resistivity;Output layer output is corresponding desired output, i.e. the rich water feelings of rock mass
Condition.
5. a kind of tunnel rich water rock mass Synthetic Geological Prediction Ahead of Construction method based on BP neural network as described in claim 1,
It is characterized in that:In the step (3), the interstitial content of input layer, hidden layer and output layer is determined, using BP neural network algorithm
Establish rich water rock mass prediction model.
6. a kind of tunnel rich water rock mass Synthetic Geological Prediction Ahead of Construction method based on BP neural network as described in claim 1,
It is characterized in that:In the step (3), the output of neural network model is rich water situation, wherein being rich in underground water assignment 1;Containing big
Measure underground water assignment 2;Containing a small amount of underground water assignment 3;Dry anhydrous assignment 0.
Specifically, geological information is as the first input factor, the wherein very big assignment 1 of rich water possibility, the larger tax of rich water possibility
Value 2, possible aqueous assignment 3, substantially anhydrous assignment 0;
Face water outlet status is as the second input factor, and wherein water drenching and strand water assignment 1, linear water assignment 2, humidity is dripped
Assignment 3, dry anhydrous assignment 0;
Vp/Vs inputs the factor as third, wherein very big assignment 1, larger assignment 2, bigger assignment 3, substantially anhydrous variation are assigned
Value 0;
Electromagnetic waveform feature is as the 4th input factor, wherein discretization strong reflection amplitude wave assignment 1, noisy mode assignment 2,
Amplitude change is little, close to continuous lineups assignment 3, waveform short arc, high frequency, lineups assignment 4 smoothly and continuously;
Change in apparent resistivity reduces larger assignment 2, assignment in a slight decrease as the 5th input factor wherein seriously reducing assignment 1
3, steady unchanged assignment 0;
Resistivity situation is as the 6th input factor, wherein 0~0.035k of resistivity Ω m assignment 1, and resistivity 0.035~
0.1k Ω m assignment 2, resistivity 0.1~0.275k Ω m assignment 3, resistivity are more than 0.275k Ω m assignment 4.
7. a kind of tunnel rich water rock mass Synthetic Geological Prediction Ahead of Construction method based on BP neural network as described in claim 1,
It is characterized in that:In the step (4), the typical case of collection research region rich water rock mass forecast builds learning sample, right later
Each connection weight assigns the random number in a section (0,1) respectively in neural network model, sets error function E, gives and calculates
Accuracy value and maximum study number M, network output is calculated according to input data.
8. a kind of tunnel rich water rock mass Synthetic Geological Prediction Ahead of Construction method based on BP neural network as described in claim 1,
It is characterized in that:In the step (4), corrected using the error term of each neuron of hidden layer and the input of each neuron of input layer
Hidden layer weights;The power of output unit is corrected using the error term of each neuron of output layer and the output of hidden layer each neuron
Value.
9. a kind of tunnel rich water rock mass Synthetic Geological Prediction Ahead of Construction method based on BP neural network as described in claim 1,
It is characterized in that:In the step (4), when being finished all samples, judge whether network error meets the requirements, when error reaches pre-
If precision or study number are more than default maximum times, then terminate algorithm;Otherwise, learn into next circle, until result reaches
Until default precision or study number are more than preset times;The neural network of rich water rock mass is established according to the weights and threshold value obtained
Prediction model.
10. a kind of tunnel rich water rock mass Synthetic Geological Prediction Ahead of Construction method based on BP neural network as described in claim 1,
It is characterized in that:In the step (4), by the richness of the water outlet status and above-mentioned acquisition of the rich water situation of design phase and face
The geophysical prospecting informations of water rock mass are input in the rich water rock mass prediction model based on BP neural network, the calculating through neural network point
Analysis, judges the country rock rich water situation of front of tunnel heading.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810030866.5A CN108416458A (en) | 2018-01-12 | 2018-01-12 | A kind of tunnel rich water rock mass Synthetic Geological Prediction Ahead of Construction method based on BP neural network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810030866.5A CN108416458A (en) | 2018-01-12 | 2018-01-12 | A kind of tunnel rich water rock mass Synthetic Geological Prediction Ahead of Construction method based on BP neural network |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108416458A true CN108416458A (en) | 2018-08-17 |
Family
ID=63125509
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810030866.5A Pending CN108416458A (en) | 2018-01-12 | 2018-01-12 | A kind of tunnel rich water rock mass Synthetic Geological Prediction Ahead of Construction method based on BP neural network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108416458A (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110989028A (en) * | 2019-11-26 | 2020-04-10 | 山东大学 | Tunnel bionic advanced geological prediction equipment and method based on artificial intelligence |
CN112130216A (en) * | 2020-08-19 | 2020-12-25 | 中国地质大学(武汉) | Geological advanced fine forecasting method based on convolutional neural network multi-geophysical prospecting method coupling |
CN112177617A (en) * | 2020-09-25 | 2021-01-05 | 中铁二十局集团有限公司 | Advanced geological forecast prediction method and system for high-pressure water-rich fault tunnel construction |
CN113010942A (en) * | 2021-02-25 | 2021-06-22 | 中国铁路设计集团有限公司 | Tunnel excavation risk early warning and surrounding rock grading evaluation method |
CN113671585A (en) * | 2021-08-18 | 2021-11-19 | 中国矿业大学 | Intelligent transient electromagnetic detection and real-time early warning method for excavation roadway |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107729716A (en) * | 2017-11-27 | 2018-02-23 | 西安建筑科技大学 | A kind of mine water inrush Forecasting Methodology based on long Memory Neural Networks in short-term |
-
2018
- 2018-01-12 CN CN201810030866.5A patent/CN108416458A/en active Pending
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107729716A (en) * | 2017-11-27 | 2018-02-23 | 西安建筑科技大学 | A kind of mine water inrush Forecasting Methodology based on long Memory Neural Networks in short-term |
Non-Patent Citations (2)
Title |
---|
廖志泓: "基于BP 神经网络模型预测隧道涌水量的探讨", 《铁路工程技术与经济》 * |
李天斌: "隧道超前地质预报综合分析方法", 《岩石力学与工程学报》 * |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110989028A (en) * | 2019-11-26 | 2020-04-10 | 山东大学 | Tunnel bionic advanced geological prediction equipment and method based on artificial intelligence |
CN112130216A (en) * | 2020-08-19 | 2020-12-25 | 中国地质大学(武汉) | Geological advanced fine forecasting method based on convolutional neural network multi-geophysical prospecting method coupling |
CN112177617A (en) * | 2020-09-25 | 2021-01-05 | 中铁二十局集团有限公司 | Advanced geological forecast prediction method and system for high-pressure water-rich fault tunnel construction |
CN113010942A (en) * | 2021-02-25 | 2021-06-22 | 中国铁路设计集团有限公司 | Tunnel excavation risk early warning and surrounding rock grading evaluation method |
CN113671585A (en) * | 2021-08-18 | 2021-11-19 | 中国矿业大学 | Intelligent transient electromagnetic detection and real-time early warning method for excavation roadway |
CN113671585B (en) * | 2021-08-18 | 2022-04-29 | 中国矿业大学 | Intelligent transient electromagnetic detection and real-time early warning method for excavation roadway |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108416458A (en) | A kind of tunnel rich water rock mass Synthetic Geological Prediction Ahead of Construction method based on BP neural network | |
CN106781489B (en) | A kind of road network trend prediction method based on recurrent neural network | |
CN108985135A (en) | A kind of human-face detector training method, device and electronic equipment | |
CN108664893A (en) | A kind of method for detecting human face and storage medium | |
CN106570516A (en) | Obstacle recognition method using convolution neural network | |
CN109800863A (en) | A kind of well-log facies recognition method based on fuzzy theory and neural network | |
CN110401978B (en) | Indoor positioning method based on neural network and particle filter multi-source fusion | |
US20090276157A1 (en) | System and method for interpretation of well data | |
CN109242149A (en) | A kind of student performance early warning method and system excavated based on educational data | |
CN106529667A (en) | Logging facies identification and analysis method based on fuzzy depth learning in big data environment | |
CN104091216A (en) | Traffic information predication method based on fruit fly optimization least-squares support vector machine | |
CN106529185B (en) | A kind of combination forecasting method and system of ancient building displacement | |
CN108376295A (en) | A kind of oil gas dessert prediction technique and storage medium | |
CN105760673A (en) | Fluvial facies reservoir earthquake sensitive parameter template analysis method | |
CN110472363B (en) | Surrounding rock deformation grade prediction method and system suitable for high-speed railway tunnel | |
CN105869100A (en) | Method for fusion and prediction of multi-field monitoring data of landslides based on big data thinking | |
CN107783185A (en) | Chromatography static correction processing method and device | |
CN106353820A (en) | Method for interpreting well logging on basis of knowledge models | |
CN112035941A (en) | Prediction method for surface subsidence of deep foundation pit excavation based on BAS-BP model | |
CN113610945A (en) | Ground stress curve prediction method based on hybrid neural network | |
CN103637800B (en) | Based on the method for the bioelectrical impedance analysis of eight sections of impedance models | |
Robati et al. | Inflation rate modeling: adaptive neuro-fuzzy inference system approach and particle swarm optimization algorithm (ANFIS-PSO) | |
KR20200058258A (en) | System and method for predicting ground layer information, and a recording medium having computer readable program for executing the method | |
CN108027896A (en) | System and method for decoding the pulse reservoir with continuous synaptic plasticity | |
CN115948964A (en) | Road flatness prediction method based on GA-BP neural network |
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 |