CN110006348A - A kind of early warning type Shield-bored tunnels segments' joints waterproofing performance intelligent monitoring method - Google Patents
A kind of early warning type Shield-bored tunnels segments' joints waterproofing performance intelligent monitoring method Download PDFInfo
- Publication number
- CN110006348A CN110006348A CN201910209178.XA CN201910209178A CN110006348A CN 110006348 A CN110006348 A CN 110006348A CN 201910209178 A CN201910209178 A CN 201910209178A CN 110006348 A CN110006348 A CN 110006348A
- Authority
- CN
- China
- Prior art keywords
- data
- seam
- optical fibre
- segments
- neural network
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B11/00—Measuring arrangements characterised by the use of optical techniques
- G01B11/02—Measuring arrangements characterised by the use of optical techniques for measuring length, width or thickness
-
- 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)
- Theoretical Computer Science (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Lining And Supports For Tunnels (AREA)
Abstract
The present invention relates to a kind of early warning type Shield-bored tunnels segments' joints waterproofing performance intelligent monitoring methods, to be monitored to segments' joints waterproofing performance, used segment joint monitoring system includes displacement sensor monitoring device, it is axially fixed to the sliding rail of tunnel duct piece inner wall, movable data receiver and host computer, wherein, displacement sensor monitoring device includes by optical fibre displacement sensor, reflector plate, laser light source, photoelectric converter, data collector, movable data receiver can be moved along sliding rail, it is received to the data periodically to the data collector being distributed in shield tunnel and is transported to host computer.
Description
Technical field
The invention belongs to shield tunnel project fields, are more particularly to a kind of Shield-bored tunnels seam waterproof performance intelligence
Monitoring method.
Background technique
During Shield-bored tunnels operation, there are unfavorable to tunnel structure health for the leakage problem of segment joint position
It influences, the durability and applicability in tunnel may be weakened, if administering not in time, the outer lateral seepage of one side pipe ring causes soil around
Grain is lost, and causes change of stress field on the outside of pipe ring, and certain adverse effect is generated to pipe ring structure, and the infiltration of another aspect seam is logical
It is penetrated into inside segment joint after crossing gasket, under certain gas concentration lwevel and relative humidity conditions, it is mixed to easily cause connector
The carbonization for coagulating soil accelerates the corrosion of reinforcing bar and bolt, reduces the rigidity of duct piece connector as carbonation depth constantly increases.Cause
This, it is essential for the monitoring of shield tunnel seam waterproof to guarantee the normal use in Shield-bored tunnels.
Domestic traditional shield tunnel percolating water monitoring method is to arrange 4 to have the personnel of professional qualification to shield at present
Tunnel carries out periodic detection, in which: illuminates, takes pictures, detecting tunnel and record 1 people of each arrangement, it is generally the case that completes 1 area
Between section need to spend time about 3h to 4h, need to arrange personnel after obtaining data and arrange and analyze, 1 shield section needs
It wants 3 to 4 position personnel to spend the 1d time, by taking 1 typical subway tunnel line, 20 sections as an example, completes the detection of 1 route
About 800h and 160 person-time is needed, in the detection process, since staff's carelessness causes the ignoring of test point, the error of data
Etc. reasons reduce manual detection efficiency, while the detection and analysis excessive cycle and personnel demand of conventional method are excessive, greatly
The detection efficiency for reducing Shield-bored tunnels waterproof performance.Therefore, it is necessary to take more efficiently shield tunnel waterproof
Performance monitoring system, to meet increasingly huge subway system waterproof monitoring requirements.
Detection with artificial intelligence in the high speed development of all trades and professions, especially in mechanical manufacturing field intelligent monitor system
Precision is continuously improved, and largely instead of artificial detection method, therefore is equally applicable to Shield-bored tunnels monitoring system
In.
Summary of the invention
The object of the present invention is to provide a kind of Shield-bored tunnels segments' joints waterproofing performance monitoring systems, on this basis
Provide the intelligent monitoring method that prediction with early warning can be accurately carried out to shield tunnel seam waterproof ability.Technical solution is such as
Under:
A kind of early warning type Shield-bored tunnels segments' joints waterproofing performance intelligent monitoring method, to segments' joints waterproofing
Performance is monitored, and used segment joint monitoring system includes displacement sensor monitoring device, is axially fixed to tunnel
Sliding rail, movable data receiver and the host computer of section of jurisdiction inner wall, wherein
Displacement sensor monitoring device includes optical fibre displacement sensor, reflector plate, laser light source, photoelectric converter, data
Collector；
The optical fibre displacement sensor includes input optical fibre and output optical fibre, and input optical fibre and output optical fibre are embedded in section of jurisdiction
The side of seam is carried out to laser conduction；
The reflector plate has the reflection bar of different specular reflectivitys including multiple groups, is laid in and connects with fibre optical sensor
The reflection band of the opposition joint face in seam face, different specular reflectivitys carries out multi-level feedback to input light intensity；
The laser light source is connect with input optical fibre, provides laser intensity for input optical fibre；
The photoelectric converter is connect with output optical fibre, converts optical signal into electric signal；
The data collector is connect with photoelectric converter, to store to shield tunnel monitoring real time data；
Movable data receiver can be moved along sliding rail, to periodically to the data collector being distributed in shield tunnel
Data received and be transported to host computer.
The monitoring method realized using above system is as follows: host computer is based on BP neural network algorithm and carries out machine learning:
By being trained to multiple segments' joints waterproofing waterproof test data samples, according to the feedforward transmitting of gradient decline and reversely
The error that learning rules reduce target output value and real output value is propagated, seam joint open and faulting of slab ends amount and waterproof ability are established
Characteristic relation, so that decision and early warning are carried out to operation phase tunnel duct piece seam monitoring data using neural network, under
Column step is monitored:
(1) acquisition and processing of data sample
It acquires data and mainly tests acquirement from the multiple groups of segment joint watertightness, obtain in different faulting of slab ends amount SiUnder the conditions of connect
Stitch waterproof performance Paver,waterWith seam joint open ΔiBetween non-linear relation, every group of test data includes following parameter: pipe
Piece seam joint open (Δi), segment joint faulting of slab ends amount (Si), gasket hardness (A), seam waterproof ability (Paver,water), it will be upper
Data are stated to be grouped as training data and test data；
The data of acquisition are normalized, it is made to be converted into the respective value in [0,1] section；
(2) foundation of BP neural network
The shield duct piece seam waterproof capability learning model based on BP neural network is established, which includes one
A input layer, a hidden layer and an output layer, wherein input layer includes 3 neurons: segment joint joint open, section of jurisdiction connect
Stitch faulting of slab ends amount, gasket hardness；Output layer includes a neural unit, i.e. segments' joints waterproofing ability；Input layer and output layer
Neuron be all made of the test sample data after step (1) normalized；
(3) neural network feedforward transmitting and test
Neural metwork training stop condition: maximum number of iterations and training objective error is set, at step (1) normalization
Test sample data after reason are input to the neural-network learning model of step (2) building, will after the completion of sample data training
Segments' joints waterproofing test data substitutes into the neural variable matrix of training and tests, and training objective error is counted using mean square deviation
It calculates；
(4) neural network output data, that is, waterproof ability numerical value backpropagation in test data；
(5) segment joint monitoring data during operation are handled using the neural network model of foundation.
Preferably, according to the non-linear behavior of input training sample, hidden layer and output layer nerve are single in the neural network
The activation primitive of member is all made of to Sigmoid type conversion function derivation；The neural-network learning model passes through in Python
Scikit-learn is realized in library.,
(5) method is as follows: the joint open Δ monitored with neural network model to tunnel duct piece seami,monitorAnd mistake
Platform amount Si,monitorIt is diagnosed and is analyzed, and export seam waterproof ability actual value, pass through discriminatory analysis:
In formula, Pneuron,waterSeam waterproof ability value, P are exported for neural network modeldesign,waterIt is anti-for segment joint
Water design value；
When 1. the condition meets, segments' joints waterproofing ability is normal；When 2. the condition meets, then segment joint
Waterproof ability failure, by positioning, counting the distribution of seam waterproof failpoint and monitor value, diagnosis tunnel water proofing failure cause is simultaneously pre-
Seam waterproof capacity variation trend is surveyed, and carries out early warning.
The present invention has the following advantages and beneficial effects: compared with traditional technology
1. reflective displacement sensor is arranged in shield duct piece seam crossing in the present invention, realizes and be based on input optical fibre and output light
Fine light intensity changes to segment joint joint open and faulting of slab ends amount real-time monitoring；
2. segment joint monitoring system of the present invention is by being arranged displacement sensor in seam crossing and data being arranged at section of jurisdiction
Collector realizes the unmanned monitoring of duct pieces of shield tunnel seam, effective to improve monitoring accuracy and efficiency；
3. seam waterproof ability intelligence learning of the present invention and decision are based on BP neural network algorithm and complete machine learning, formed
Neural network model with waterproof ability evaluating ability, neuron models are diagnosed and are divided to segment joint monitoring data
The intelligence of seam waterproof merit rating, the effective service life for extending Shield-bored tunnels are realized in analysis；
Detailed description of the invention
Fig. 1 is Shield-bored tunnels segments' joints waterproofing ability intelligent monitoring of the present invention and early warning flow chart；
Fig. 2 is that Shield-bored tunnels segment joint of the present invention monitors system schematic；
Fig. 3 is the reflective displacement sensor schematic illustration of the present invention；
Fig. 4 is the reflective displacement sensor layout diagram of the present invention；
Fig. 5 is the optical fiber micro-displacement sensor simulation curve figure of reflective light intensity under the conditions of different reflectivity；
Fig. 6 is neural-network learning model schematic diagram of the present invention.
Include in Fig. 1~6:
1- Shield-bored tunnels；
2- tunnel vault；
The tunnel 3- haunch；
Encircle bottom in the tunnel 4-；
The tunnel 5- longitudinal joint；
The tunnel 6- circumferential weld；
7- data collector；
8- sliding rail；
9- movable data receiver；
10- optical fibre displacement sensor；
11- input optical fibre；
12- output optical fibre；
13- self-adhesive type reflector plate；
14- reflection bar；
15- rubber gasket；
On the outside of 16- seal groove；
On the inside of 17- seal groove；
18- laser light source；
19- photoelectric converter.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, below in conjunction with attached drawing, to the present invention into
Row is further described:
The present invention is based on BP artificial neural network intelligent algorithms, for Shield-bored tunnels seam waterproof traditional detection method
The problem, the formula that provides alert Shield-bored tunnels segments' joints waterproofing performance intelligent monitor system, by being opened with seam
The sample data of amount and faulting of slab ends scale sign seam waterproof ability exercises supervision study, formation seam waterproof ability neural network mould
Type is diagnosed and is analyzed to molding duct pieces of shield tunnel seam monitoring joint open and faulting of slab ends amount, in this, as intermediate bridge, done
The prediction of segments' joints waterproofing ability and early warning out.
It is made of referring to Shield-bored tunnels segments' joints waterproofing ability intellectual faculties described in Fig. 13 parts, respectively
System, seam waterproof ability intelligent learning system, the prediction of seam waterproof ability and early warning system are monitored for segment joint.
It includes displacement sensor monitoring device, the cunning for being axially fixed to tunnel duct piece inner wall that segment joint, which monitors system,
Rail, movable data receiver and host computer, wherein displacement sensor monitoring device includes optical fibre displacement sensor, self-adhesive type
Reflector plate, laser light source, photoelectric converter, data collector.
1 segment joint of Shield-bored tunnels is divided into tunnel longitudinal joint 5 between section of jurisdiction and section of jurisdiction and pipe ring and pipe referring to fig. 2
The tunnel circumferential weld 6 of interannular, tunnel longitudinal joint 5 and tunnel circumferential weld 6 are the main portions of shield tunnel waterproof, are shown according to existing research
Tunnel vault 2, tunnel haunch 3 and the tunnel arch segment joint during operation of bottom 4 of Shield-bored tunnels 1 are easy to happen infiltration
Leakage, typically resulting in seam the factor of seepage flow occurs is seam joint open ΔiExcessive and faulting of slab ends amount SiIt is excessive, segment joint joint open
It is closely related with faulting of slab ends amount and gasket contact stress, thus the monitoring system mainly tunnel vault 2, tunnel haunch 3 with
And tunnel arch bottom 4 lay reflective optical fiber displacement sensor (being not drawn into figure), with run the joint open exported in phase time domain and
The variation of faulting of slab ends amount carries out characterization seam waterproof capacity variation, and data collector 7 is displaced the reflection type optical fiber at segment joint and passes
The monitoring data of sensor are received, stored and are transmitted, and movable data receiver 9 is along sliding rail 8 periodically in shield tunnel
The data of the data collector 7 of distribution are received and are transported to host computer.
Referring to seam joint open Δ described in Fig. 3iWith faulting of slab ends amount SiMonitoring device is mainly pasted by optical fibre displacement sensor 10 and certainly
Formula reflector plate 13 forms, and optical fibre displacement sensor 10 and self-adhesive type reflector plate 13 are located at segment joint face two sides, optical fiber position
It include input optical fibre 11 and output optical fibre 12 in displacement sensor 10, light enters in input optical fibre 11 from light source couples, through self-adhesive type
Reflector plate 13 is reflected onto output optical fibre 12 again, due to the light intensity difference of input optical fibre 11 and output optical fibre 12, is based on this
Principle measure to segment joint, sets 11 fibre of input optical fibre and brings out and penetrates optical field distribution as axial symmetry Gaussian Profile mould
Type receives light intensity with output optical fibre 12 to 11 output intensity of input optical fibre and derives are as follows:
In formula, I (Δi) it is that output optical fibre 12 receives light intensity, I0For 11 output intensity of input optical fibre, a is fiber radius, Ri
For the specular reflectivity of self-adhesive type reflector plate 13, ΔiFor segment joint joint open, θcFor angle of reflection.
The self-adhesive type reflector plate 13 is made of different groups of reflection bars 14, and every group of reflector plate calculating parameter is distinguished from top to bottom
Are as follows:
[R-i,H-i],……,[R-1,H-1], [R0,H0], [R1,H1],……,[Ri,Hi]
Wherein RiThe specular reflectivity of lower respectively different number grade, H to be aboveiFor the width of reflection bar 14, due to amount
Survey seam faulting of slab ends amount magnitude is 1mm, therefore H-i=...=H-1=H0=...=Hi=1mm passes through different mirror-reflections
Rate sets different seam faulting of slab ends values.
When the output optical fibre 12 is located within reflective taper, reflected light is received by output optical fibre 12, and is obtained not
Same RiUnder the conditions of I (Δi)~ΔiRelation curve.Faulting of slab ends amount S between seamiBy the specular reflectivity R of different number gradeiIt calculates
Curvilinear characteristic judged, the joint open Δ between seamiBy specular reflectivity RiUnder the conditions of I (Δi)~ΔiCurve can obtain
Out.
Segments' joints waterproofing ability is mainly compressed by rubber gasket 15 and is realized referring to fig. 4, when shield tunnel is in deep ground
When in layer, the reflective optical fiber displacement sensor at tunnel vault 2 and tunnel arch bottom 4 is mainly disposed to 17 on the inside of seal groove, tunnel
The reflective optical fiber displacement sensor of haunch 3 is mainly disposed to 16 on the outside of seal groove.When shield tunnel is in shallow stratum, tunnel
Road vault 2 and the reflective optical fiber displacement sensor at tunnel arch bottom 4 are mainly disposed to 16 on the outside of seal groove, tunnel haunch 3 it is anti-
The formula optical fibre displacement sensor of penetrating is mainly disposed to 17 on the inside of seal groove.The optical fibre displacement sensor of reflective optical fiber displacement sensor
10 are embedded in section of jurisdiction, and fine end is flushed with joint face, and self-adhesive type reflector plate 13 is labelled to joint surface, 10 optical fiber edge of fibre optical sensor
Segment joint face is laid in parallel, the external laser light source 18 of input optical fibre 11, the external photoelectric converter 19 of output optical fibre 12, optical signal
By light intensity and joint open Δ after photoelectric converter 19iMapping relations be converted to voltage and joint open ΔiMapping relations,
The electric signal that data collector 7 exports photoelectric converter 19 is collected and stores.
In conjunction with the embodiments by the reflective optical fiber displacement sensor, the present invention is described in more detail.Utilize I
(Δi) indicate optical fiber output characteristic modulation function, showed using following formula to optical fiber output light intensity affecting parameters:
I(Δi)=f (a, Ri,θc)
Simulate different specular reflectivitys to method using in optic fiber displacement sensor using MATLAB software control quantity method numerical simulation
The influence of device obtains the optical fiber micro-displacement sensor theoretical curve of reflective light intensity under the conditions of different reflectivity, as shown in Figure 5.Its
In, 11 output intensity I of input optical fibre0=60 × 108Cd, input optical fibre 11 and output optical fibre 12 radius a=0.2mm, angle of reflection θc
=15 °, 13 specular reflectivity R of self-adhesive type reflector plateiAccording to different faulting of slab ends amount SiIt is demarcated, reflection bar 14 is taken as 0.1 respectively,
0.2 ..., 0.9,1.0 etc. 10 reflectivity.
BP neural network algorithm, which is based primarily upon, referring to seam waterproof ability intelligent learning system described in Fig. 6 carries out engineering
It practises, by being trained to a large amount of segments' joints waterproofing waterproof test data sample, is transmitted according to the feedforward of gradient decline
The error for reducing target output value and real output value with back-propagation learning rule, effectively establishes seam joint open and faulting of slab ends
The characteristic relation of amount and waterproof ability, so as to carry out decision to operation phase tunnel duct piece seam monitoring data using neural network
With early warning.
It is as follows that algorithm learns key step:
(1) acquisition and processing of data sample
It acquires data and mainly tests acquirement from the multiple groups of segment joint watertightness, content of the test is in different faulting of slab ends amount Si
Under the conditions of seam waterproof performance Paver,waterWith seam joint open ΔiBetween non-linear relation, therefore, every group of test data packet
Include following parameter: segment joint joint open (Δi), segment joint faulting of slab ends amount (Si), gasket hardness (A), seam waterproof ability
(Paver,water), above-mentioned data are grouped as training data and test data.
Since the order of magnitude great disparity of parameters need to carry out the data of acquisition for the convergence rate for accelerating neural network
Normalized makes it be converted into the respective value in [0,1] sectionIts calculation formula is:
In formula, XminFor the minimum value of each parameter, XmaxFor the maximum value of each parameter, XiFor the acquisition data of each parameter,
For each parameter normalization treated numerical value.
(2) foundation of BP neural network
The shield duct piece seam waterproof capability learning model based on BP neural network is established, which includes one
A input layer, a hidden layer and an output layer, wherein input layer includes 3 neurons: segment joint joint open, section of jurisdiction connect
Stitch faulting of slab ends amount, gasket hardness；Output layer includes a neural unit, i.e. segments' joints waterproofing ability；Hidden layer node number
It can be obtained by empirical equation:
In formula, m is input layer number, and n is output layer node number, and a is the regulating constant between 1~10.According to
Can be calculated hidden layer node number in neural network is 6.
According to the non-linear behavior of input training sample, the activation of hidden layer and output layer neural unit in the neural network
Function is all made of to Sigmoid type conversion function derivation, formula are as follows:
The neural-network learning model can be realized in library by scikit-learn in Python, input layer and output
The neuron of layer is all made of the test sample data after step (1) normalized.
(3) neural network feedforward transmitting and test
Neural metwork training stop condition: maximum number of iterations and training objective error is set, at step (1) normalization
Test sample data after reason are input to the neural-network learning model of step (2) building, will after the completion of sample data training
Segments' joints waterproofing test data substitutes into the neural variable matrix of training and tests.Wherein, the training objective error is using equal
Fang Jinhang is calculated, formula are as follows:
In formula, Q is input layer set, and Y (k) is neural network prediction value, and t (k) is neural network output
Layer real output value, makes mean square error be less than training objective error.
(4) neural network output data, that is, waterproof ability numerical value backpropagation in test data
After neural network test is completed, output data is subjected to anti-normalization processing, the process flow:
In formula, Y is the output valve after anti-normalization processing, YminFor segments' joints waterproofing ability in waterproof test data
Minimum value, YmaxFor the maximum value of segments' joints waterproofing ability in waterproof test data,It is exported for neural network practical
Value.
Compare the output valve of waterproof test data Yu neural network model renormalization, evaluates training neural network learning
The accuracy and applicability of model.
(5) segment joint monitoring data during operation are handled using the neural network model of foundation
The joint open Δ that tunnel duct piece seam is monitored with the neural network model with applicabilityi,monitorAnd faulting of slab ends
Measure Si,monitorIt is diagnosed and is analyzed, and export seam waterproof ability actual value, pass through discriminatory analysis:
In formula, Pneuron,waterSeam waterproof ability value, P are exported for neural network modeldesign,waterIt is anti-for segment joint
Water design value.
When 1. the condition meets, segments' joints waterproofing ability is normal；When 2. the condition meets, then segment joint
Waterproof ability failure, by positioning, counting the distribution of seam waterproof failpoint and monitor value, diagnosis tunnel water proofing failure cause is simultaneously pre-
Seam waterproof capacity variation trend is surveyed, early warning is carried out to the analysis and prediction, and propose related disposing suggestion.
The foregoing is merely the preferred embodiment of the present invention, all technical solutions belonged under thinking of the present invention belong to this
The protection scope of invention.It should be pointed out that all within the spirits and principles of the present invention, made any modification replaces on an equal basis, changes
Into etc., it should all be included in the protection scope of the present invention.
Claims (3)
1. a kind of early warning type Shield-bored tunnels segments' joints waterproofing performance intelligent monitoring method, to segments' joints waterproofing
It can be carried out monitoring, used segment joint monitoring system includes displacement sensor monitoring device, is axially fixed to tunneltron
Sliding rail, movable data receiver and the host computer of piece inner wall, wherein
Displacement sensor monitoring device includes optical fibre displacement sensor, reflector plate, laser light source, photoelectric converter and data acquisition
Device；
The optical fibre displacement sensor includes input optical fibre and output optical fibre, and input optical fibre and output optical fibre are embedded in segment joint
Side, carry out to laser conduction；
The reflector plate has the reflection bar of different specular reflectivitys including multiple groups, is laid in fibre optical sensor joint face
Opposition joint face, the reflection bands of different specular reflectivitys carries out multi-level feedback to input light intensity；
The laser light source is connect with input optical fibre, provides laser intensity for input optical fibre；
The photoelectric converter is connect with output optical fibre, converts optical signal into electric signal；
The data collector is connect with photoelectric converter, to store to shield tunnel monitoring real time data；
Movable data receiver can be moved along sliding rail, to the data to the data collector being distributed in shield tunnel into
Row receives and is transported to host computer；
It can be such that host computer is based on BP neural network algorithm and carries out machine learning using the monitoring method that above system is realized:
By being trained to multiple segments' joints waterproofing waterproof test data samples, according to the feedforward transmitting of gradient decline and reversely
The error that learning rules reduce target output value and real output value is propagated, seam joint open and faulting of slab ends amount and waterproof ability are established
Characteristic relation, so that decision and early warning are carried out to operation phase tunnel duct piece seam monitoring data using neural network, under
Column step is monitored:
(1) acquisition and processing of data sample
It acquires data and mainly tests acquirement from the multiple groups of segment joint watertightness, obtain in different faulting of slab ends amount SiUnder the conditions of seam it is anti-
Aqueous energy Paver,waterWith seam joint open ΔiBetween non-linear relation, every group of test data includes following parameter: section of jurisdiction connects
Stitch joint open (Δi), segment joint faulting of slab ends amount (Si), gasket hardness (A), seam waterproof ability (Paver,water), by above-mentioned number
According to being grouped as training data and test data；
The data of acquisition are normalized, it is made to be converted into the respective value in [0,1] section；
(2) foundation of BP neural network
The shield duct piece seam waterproof capability learning model based on BP neural network is established, which includes one defeated
Enter layer, a hidden layer and an output layer, wherein input layer includes 3 neurons: segment joint joint open, segment joint are wrong
Platform amount, gasket hardness；Output layer includes a neural unit, i.e. segments' joints waterproofing ability；The mind of input layer and output layer
Test sample data after member is all made of step (1) normalized；
(3) neural network feedforward transmitting and test
Neural metwork training stop condition: maximum number of iterations and training objective error is set, after step (1) normalized
Test sample data be input to step (2) building neural-network learning model, sample data training after the completion of, by section of jurisdiction
Seam waterproof test data substitutes into the neural variable matrix of training and tests, and training objective error is calculated using mean square deviation；
(4) neural network output data, that is, waterproof ability numerical value backpropagation in test data；
(5) segment joint monitoring data during operation are handled using the neural network model of foundation.
2. the method according to claim 1, wherein according to input training sample non-linear behavior, the nerve
The activation primitive of hidden layer and output layer neural unit is all made of to Sigmoid type conversion function derivation in network；The nerve net
Network learning model is realized in library by scikit-learn in Python.
3. the method according to claim 1, wherein (5) execute in following manner: using neural network model
To the joint open Δ of tunnel duct piece seam monitoringi,monitorWith faulting of slab ends amount Si,monitorIt is diagnosed and is analyzed, and it is anti-to export seam
Outlet capacity actual value, passes through discriminatory analysis:
In formula, Pneuron,waterSeam waterproof ability value, P are exported for neural network modeldesign,waterIt is set for segments' joints waterproofing
Evaluation；
When 1. the condition meets, segments' joints waterproofing ability is normal；When 2. the condition meets, then segments' joints waterproofing
Ability failure, by positioning, counting the distribution of seam waterproof failpoint and monitor value, diagnosing tunnel water proofing failure cause and predicting to connect
Waterproof ability variation tendency is stitched, and carries out early warning.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910209178.XA CN110006348B (en) | 2019-03-19 | 2019-03-19 | Intelligent monitoring method for waterproof performance of segment joint of early warning type subway shield tunnel |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910209178.XA CN110006348B (en) | 2019-03-19 | 2019-03-19 | Intelligent monitoring method for waterproof performance of segment joint of early warning type subway shield tunnel |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110006348A true CN110006348A (en) | 2019-07-12 |
| CN110006348B CN110006348B (en) | 2021-05-11 |
Family
ID=67167398
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910209178.XA Active CN110006348B (en) | 2019-03-19 | 2019-03-19 | Intelligent monitoring method for waterproof performance of segment joint of early warning type subway shield tunnel |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110006348B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112324457A (en) * | 2020-10-16 | 2021-02-05 | 中铁十四局集团有限公司 | Water-rich super-large section shield tunnel waterproof composite lining structure and construction method |
| CN112324456A (en) * | 2020-10-16 | 2021-02-05 | 中铁十四局集团有限公司 | Shield tunnel composite lining structure and method |
| CN113569362A (en) * | 2021-09-22 | 2021-10-29 | 南昌市建筑科学研究所（南昌市建筑工程质量检测中心） | System and method for managing whole life cycle of shield segment |
| CN114518203A (en) * | 2022-01-26 | 2022-05-20 | 中交隧道工程局有限公司 | Shield tunnel segment leakage point identification method |
| CN114858079A (en) * | 2022-05-23 | 2022-08-05 | 青岛益群地下城开发有限公司 | Tunnel deformation monitoring method based on distributed multipoint strain and displacement conversion network |
| CN115711706A (en) * | 2023-01-09 | 2023-02-24 | 中南大学 | Tunnel joint sealing gasket waterproof capacity early warning system and judgment method |
| CN116989680A (en) * | 2023-09-28 | 2023-11-03 | 武汉科技大学 | Fiber bragg grating monitoring device and method for joints in shield segment assembly process |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH0673997A (en) * | 1992-08-25 | 1994-03-15 | Shimizu Corp | Construction of waterway tunnel |
| CN202690083U (en) * | 2012-01-13 | 2013-01-23 | 上海市政工程设计研究总院（集团）有限公司 | Waterproof structure of shield tunnel duct piece |
| US20140060199A1 (en) * | 2012-09-05 | 2014-03-06 | Qorex Llc | Multi-mode holographic pressure sensor |
| CN104240470A (en) * | 2013-06-09 | 2014-12-24 | 同济大学 | Wireless joints displacement measuring system for metro shield tunnel |
| CN105987778A (en) * | 2015-05-06 | 2016-10-05 | 中铁建设投资集团有限公司 | Shield tunnel segment joint stress in-situ measurement method |
| CN207033473U (en) * | 2017-07-17 | 2018-02-23 | 中铁隧道勘测设计院有限公司 | A kind of waterproof structure of shield tunnel segment seam |
| CN108489456A (en) * | 2018-03-30 | 2018-09-04 | 胡仲春 | tunnel monitoring system and tunnel monitoring method |
| CN109387385A (en) * | 2018-10-24 | 2019-02-26 | 中国科学院武汉岩土力学研究所 | A kind of three-dimensional TBM shield outer surface country rock Load Identification Methods |
-
2019
- 2019-03-19 CN CN201910209178.XA patent/CN110006348B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH0673997A (en) * | 1992-08-25 | 1994-03-15 | Shimizu Corp | Construction of waterway tunnel |
| CN202690083U (en) * | 2012-01-13 | 2013-01-23 | 上海市政工程设计研究总院（集团）有限公司 | Waterproof structure of shield tunnel duct piece |
| US20140060199A1 (en) * | 2012-09-05 | 2014-03-06 | Qorex Llc | Multi-mode holographic pressure sensor |
| CN104240470A (en) * | 2013-06-09 | 2014-12-24 | 同济大学 | Wireless joints displacement measuring system for metro shield tunnel |
| CN105987778A (en) * | 2015-05-06 | 2016-10-05 | 中铁建设投资集团有限公司 | Shield tunnel segment joint stress in-situ measurement method |
| CN207033473U (en) * | 2017-07-17 | 2018-02-23 | 中铁隧道勘测设计院有限公司 | A kind of waterproof structure of shield tunnel segment seam |
| CN108489456A (en) * | 2018-03-30 | 2018-09-04 | 胡仲春 | tunnel monitoring system and tunnel monitoring method |
| CN109387385A (en) * | 2018-10-24 | 2019-02-26 | 中国科学院武汉岩土力学研究所 | A kind of three-dimensional TBM shield outer surface country rock Load Identification Methods |
Non-Patent Citations (1)
| Title |
|---|
| 王飞等: "基于BOTDA光纤传感技术的盾构隧道变形感知方法", 《岩石力学与工程学报》 * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112324457A (en) * | 2020-10-16 | 2021-02-05 | 中铁十四局集团有限公司 | Water-rich super-large section shield tunnel waterproof composite lining structure and construction method |
| CN112324456A (en) * | 2020-10-16 | 2021-02-05 | 中铁十四局集团有限公司 | Shield tunnel composite lining structure and method |
| CN112324457B (en) * | 2020-10-16 | 2021-08-03 | 中铁十四局集团有限公司 | Water-rich super-large section shield tunnel waterproof composite lining structure and construction method |
| CN112324456B (en) * | 2020-10-16 | 2022-02-08 | 中铁十四局集团有限公司 | Shield tunnel composite lining structure and method |
| CN113569362A (en) * | 2021-09-22 | 2021-10-29 | 南昌市建筑科学研究所（南昌市建筑工程质量检测中心） | System and method for managing whole life cycle of shield segment |
| CN113569362B (en) * | 2021-09-22 | 2021-12-21 | 南昌市建筑科学研究所（南昌市建筑工程质量检测中心） | System and method for managing whole life cycle of shield segment |
| CN114518203A (en) * | 2022-01-26 | 2022-05-20 | 中交隧道工程局有限公司 | Shield tunnel segment leakage point identification method |
| CN114518203B (en) * | 2022-01-26 | 2024-02-20 | 中交隧道工程局有限公司 | Shield tunnel segment leakage point identification method |
| CN114858079A (en) * | 2022-05-23 | 2022-08-05 | 青岛益群地下城开发有限公司 | Tunnel deformation monitoring method based on distributed multipoint strain and displacement conversion network |
| CN114858079B (en) * | 2022-05-23 | 2022-11-25 | 青岛益群地下城开发有限公司 | Tunnel deformation monitoring method based on distributed multipoint strain and displacement conversion network |
| CN115711706A (en) * | 2023-01-09 | 2023-02-24 | 中南大学 | Tunnel joint sealing gasket waterproof capacity early warning system and judgment method |
| CN116989680A (en) * | 2023-09-28 | 2023-11-03 | 武汉科技大学 | Fiber bragg grating monitoring device and method for joints in shield segment assembly process |
| CN116989680B (en) * | 2023-09-28 | 2023-12-29 | 武汉科技大学 | Fiber bragg grating monitoring device and method for joints in shield segment assembly process |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110006348B (en) | 2021-05-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110006348A (en) | A kind of early warning type Shield-bored tunnels segments' joints waterproofing performance intelligent monitoring method | |
| CN110044268A (en) | Shield tunnel seam based on fiber reflection principle, which is opened, monitors system with faulting of slab ends | |
| CN114352947B (en) | Gas pipeline leakage detection method, system, device and storage medium | |
| Ma et al. | Advances in corrosion growth modeling for oil and gas pipelines: A review | |
| Fu et al. | A hybrid deep learning approach for dynamic attitude and position prediction in tunnel construction considering spatio-temporal patterns | |
| CN100470416C (en) | Power plant thermal equipment intelligent state diagnosing and analyzing system | |
| AU2020102518A4 (en) | A method of gas sensor calibration based on linear optimization | |
| CN112906288A (en) | Switch cabinet state evaluation method based on normal sample data | |
| CN113128709B (en) | Water supply network leakage management system | |
| Wang et al. | The replacement of dysfunctional sensors based on the digital twin method during the cutter suction dredger construction process | |
| CN116027146A (en) | Multi-state distributed passive sensing monitoring method, device and equipment for distribution cable network | |
| CN111445165A (en) | Tunnel structure health monitoring online grading early warning evaluation method | |
| CN114783143A (en) | Underground cable pipe gallery construction safety monitoring and early warning analysis system based on visualization | |
| CN115654381A (en) | Water supply pipeline leakage detection method based on graph neural network | |
| CN111191855B (en) | Water quality abnormal event identification and early warning method based on pipe network multi-element water quality time sequence data | |
| CN114739891A (en) | Floating type offshore wind power mooring system corrosion state detection and evaluation system | |
| Li et al. | Deep learning and heterogeneous signal fusion approach to precursor feature recognition and early warning of coal and gas outburst | |
| Yang et al. | Method and Validation of Coal Mine Gas Concentration Prediction by Integrating PSO Algorithm and LSTM Network | |
| CN116822352A (en) | GRNN neural network-based underground waterproof sealing wall working condition prediction method and system | |
| CN114413832B (en) | Road monitoring system and method based on optical fiber sensing | |
| CN109858707A (en) | A kind of Ethylene Cracking Furnace Tubes intelligence coking diagnostic method based on ABC-ANFIS-CTF | |
| CN115563864A (en) | Intelligent prediction method for cable tunnel water leakage | |
| Zhu et al. | Prediction of distributed optical fiber monitoring data based on GRU-BP | |
| CN114332515A (en) | Online evaluation method for operation safety of coal mill based on intuition fuzzy clustering | |
| Bi et al. | A fault diagnosis algorithm for wind turbine blades based on 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |