CN109726463A - A kind of shield TBM fault early warning method based on SVM algorithm - Google Patents
A kind of shield TBM fault early warning method based on SVM algorithm Download PDFInfo
- Publication number
- CN109726463A CN109726463A CN201811589422.1A CN201811589422A CN109726463A CN 109726463 A CN109726463 A CN 109726463A CN 201811589422 A CN201811589422 A CN 201811589422A CN 109726463 A CN109726463 A CN 109726463A
- Authority
- CN
- China
- Prior art keywords
- shield
- svm
- shield tbm
- fault
- tbm
- 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
- 238000000034 method Methods 0.000 title claims abstract description 38
- 238000004422 calculation algorithm Methods 0.000 title claims abstract description 21
- 238000010276 construction Methods 0.000 claims abstract description 22
- 238000007635 classification algorithm Methods 0.000 claims abstract description 7
- 238000010801 machine learning Methods 0.000 claims abstract description 5
- 238000012706 support-vector machine Methods 0.000 claims description 52
- 230000006870 function Effects 0.000 claims description 32
- 238000012549 training Methods 0.000 claims description 13
- 230000008569 process Effects 0.000 claims description 8
- 230000001186 cumulative effect Effects 0.000 claims description 7
- 238000012360 testing method Methods 0.000 claims description 6
- 238000004458 analytical method Methods 0.000 claims description 2
- 239000002689 soil Substances 0.000 claims description 2
- 230000002159 abnormal effect Effects 0.000 abstract description 2
- 230000008901 benefit Effects 0.000 abstract description 2
- 230000034994 death Effects 0.000 abstract description 2
- 231100000517 death Toxicity 0.000 abstract description 2
- 230000009467 reduction Effects 0.000 abstract description 2
- 238000013507 mapping Methods 0.000 description 4
- 238000007726 management method Methods 0.000 description 3
- 230000005641 tunneling Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 238000009412 basement excavation Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000010224 classification analysis Methods 0.000 description 1
- 238000013499 data model Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000011067 equilibration Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 238000000611 regression analysis Methods 0.000 description 1
- 230000008521 reorganization Effects 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of shield TBM fault early warning method based on SVM algorithm, it mainly contains shield TBM fault modeling, shield TBM fault identification machine learning classification algorithm SVM modeling, trained and Classification and Identification, shield TBM fault early warning method three parts.The model of shield TBM machine learning classification algorithm SVM is trained by analyzing shield-tunneling construction sample data using this method, early warning can be carried out to equipment, the construction abnormal conditions occurred in shield TBM work progress in conjunction with selected shield TBM fault model, reduce equipment, the damage of component and the injures and deaths of personnel, and the generation of great construction accident, the security risk of construction personnel and economic loss in reduction work progress, shield industry construction information, intelligentized level are improved, there is very big economic and social benefit to shield-tunneling construction industry.
Description
Technical field
The present invention relates to tunneling shield excavation work fields, more particularly to a kind of shield TBM failure based on SVM algorithm
Method for early warning.
Background technique
Currently, the shield-tunneling construction in China is still within the fast-developing phase, the digital intelligent method of site operation also compares
Compared with shortcoming, tunneling shield tunneling construction relies primarily on artificial experience, and shield-tunneling construction project is growing day by day and shield driver quantity
Increased dramatically leads to the horizontal relative reduction of the working of shield driver, in addition geological conditions is relative complex in shield TBM construction, generates
There is a large amount of major accident and casualties in shield TBM construction at present, is badly in need of a kind of shield TBM fault early warning method at this time
That great dangerous situation early warning is reduced the generation of major accident, protects the Loss of Life and property of name, while improving shield
The digitization and intelligent level of TBM driving.
Summary of the invention
In order to solve the above technical problems, one technical scheme adopted by the invention is that a kind of shield TBM based on SVM algorithm
Fault early warning method, mainly contains shield TBM fault modeling, and shield TBM fault identification machine learning classification algorithm SVM is built
Mould, training and Classification and Identification, shield TBM fault early warning method three parts.
The technical scheme of the present invention is realized as follows: a kind of shield TBM fault early warning method based on SVM algorithm, it should
Specific step is as follows for method:
(1) shield TBM key boring parameter is created according to previous generation problem in shield TBM work progress
Shield TBM fault model；
(2) shield TBM machine learning classification algorithm SVM modeled by analysis selection shield-tunneling construction data sample, trained
It is verified with classification；
(3) Classification and Identification is carried out to shield TBM construction real time data using the SVM model after training, and fault message is carried out
Determine and pushes.
Wherein, the shield TBM key boring parameter in the step (1) refer to cutterhead revolving speed, cutter head torque, gross thrust,
Two or more combination in the shields TBM key boring parameter such as fltting speed, soil pressure.
Wherein, the detailed process of the step (2) is: preparing sample according to the shield TBM fault model that step (1) is created
Input/output argument of the notebook data as SVM, first according to one SVM of number syncaryon function creation of input variable support to
Then amount machine adjusts the parameter of kernel function, using SMO learning algorithm to SVM support vector machines and sample input/output argument into
Row training, obtain categorised decision function i.e. final SVM model, finally using final SVM model to prepare sample into
Row test verifying.
Wherein, the detailed process of the step (3) is: the SVM model pair after the test verifying obtained using step (3)
Shield TBM construction real time data carries out Classification and Identification, if being identified as fault point, stores record and cumulative time, if tired
It is more than threshold value between timing, then it is assumed that then failure pushes fault pre-alarming information to related management personnel.
Wherein, the kernel function be linear kernel function, Polynomial kernel function, gaussian kernel function, Laplce's kernel function,
Any one in sigmod kernel function.
The principle of the present invention and specific practice are: (1) shield TBM fault modeling, occur to construct in shield TBM according to previous
The problem carries out fault modeling to shield TBM key boring parameter in the process.Cutter in Mr. Yu's work progress is seriously ground
Failure is damaged, (principle only enumerates knife as a shield TBM fault model in such cases for selection cutter head torque and fltting speed
Data in disk torque and fltting speed 2), while choosing the cutter head torque and fltting speed of each time point of fault time section
Data as initial data.
(2) shield TBM failed machines learning classification algorithm SVM support vector machines
SVM (Support Vector Machine) refers to support vector machines, is a kind of common method of discrimination.In engineering
Habit field is the learning model for having supervision, commonly used to carry out pattern-recognition, classification and regression analysis.
SVM method is that sample space is mapped to a higher-dimension or even infinite dimensional feature by a Nonlinear Mapping p
In space (space Hilbert), so that being converted into feature space the problem of Nonlinear separability in original sample space
Linear separability the problem of.SVM method: using the expansion theorem of kernel function, there is no need to know the explicit table of Nonlinear Mapping
Up to formula；Due to being to establish linear learning machine in high-dimensional feature space, so not only hardly increasing meter compared with linear model
The complexity of calculation, and " dimension disaster " is avoided to a certain extent.
Prepare input/output argument of the sample data as SVM according to shield TBM fault model, this input parameter is to push away
Into speed and cutter head torque, output parameter is classification (failure, normal).
A SVM support vector machines is created according to the number combination gaussian kernel function of input variable.The number of input variable
It is 2, common kernel function has following 5 kinds:
1, linear kernel function:
2, Polynomial kernel function:
3, diameter is as base kernel function/gaussian kernel function:
4, Laplce's kernel function:
5, sigmod kernel function:
It is selected according to data training, it is most suitable for gaussian kernel function (present principles explanation is only chosen one of).Adjustment is high
The parameter of this kernel function is trained SVM support vector machines and sample input/output argument using SMO learning algorithm, obtains
Categorised decision function i.e. final SVM model.Test verifying is carried out to sample is prepared using final SVM model.
(3) shield TBM fault early warning method
Shield TBM real time data is screened by fault model, as the input data of the SVM model after training, after training
SVM category of model identification, export the type (normal point/fault point) of this group of data, extracted if being identified as normal point
Next group of real time data, record and cumulative time if being identified as fault point.Cumulative time is more than that threshold value thinks failure, is pushed away
Send warning message to related management personnel.
The beneficial effects of the present invention are: the present invention is by using a kind of fault pre-alarming side shield TBM based on SVM algorithm
Method can carry out early warning to equipment, the construction abnormal conditions occurred in shield TBM work progress, reduce equipment, component
Damage and personnel injures and deaths, and reduce work progress in the generation of great construction accident, the security risk of construction personnel and
Economic loss improves shield industry construction information, intelligentized level, has very big economy and society to shield-tunneling construction industry
It can benefit.
Detailed description of the invention
Fig. 1 is cutter head torque and fltting speed relationship shown in a kind of shield TBM fault early warning method based on SVM algorithm
Curve graph.
Fig. 2 is sample data scatter plot shown in a kind of shield TBM fault early warning method based on SVM algorithm.
Fig. 3 is a kind of training mapping graph of SVM sample data shown in the shield TBM fault early warning method based on SVM algorithm.
Fig. 4 is a kind of support of the training of SVM sample data shown in the shield TBM fault early warning method based on SVM algorithm
Vector.
Fig. 5 is a kind of test of SVM sample data model shown in the shield TBM fault early warning method based on SVM algorithm.
Fig. 6 is shield TBM fault pre-alarming flow chart.
Specific embodiment
The present invention will be further explained below with reference to the attached drawings and specific examples.
Embodiment:
Engineering introduction: the Nanjing channel Guo Jiang is that important city that " Nanjing overall city planning " determines is crossed river and quickly led to
Road connects area, Hexi new city middle bar Pukou District.Entire engineering path overall length 5853m, double hole two-wire Six Lane Designs are adopted
With " left branch of a river shield tunnel+right branch of a river bridge " scheme, wherein in left branch of a river shield tunnel (being divided into two tunnels of left and right line) north of the Changjiang River starting point
Journey is K3+600, and it is K6+532.756 that Jiangnan plum continent, which reaches mileage, and shield siding-to-siding block length is 2932.756m, two, left and right line
Two (S349, S350) Φ 14.93m mud water pressurization equilibration shield machines that German extra large Rake company production is respectively adopted in tunnel are applied
Work.
(1) shield TBM fault modeling
The equipment causes cutter to be seriously worn since geology changes plus faulty operation in the construction process.This chooses knife
As a shield TBM fault model in such cases, entity relationship diagram is as shown in Figure 1 for disk torque and fltting speed.
Blue curve is fltting speed curve, and it is fltting speed in black curve frame that red curve, which is cutter head torque curve,
Bust, the high section of torque.Fig. 1 finds out that inverse correlation obviously has occurred in torque and speed in progradation, simply
Judgement be not cutter it is out of joint be exactly that geology is mutated, on inspection for geologic change in the case where, faulty operation causes
Tool wear is serious.After geology is hardened, driving speed decline, driver increases torque, and driving speed does not still increase, hard to accelerate to lead
Cause tool wear serious.
Gross thrust and fltting speed modeling are chosen, under geological conditions situation of change, blindly driving situation tunnel different
The faulty operation of shield driver is reminded in often variation early warning in time, reduces the probability of equipment and parts damages.
(2) shield TBM failed machines learning classification algorithm SVM support vector machines
Model parameter chooses cutter head torque and fltting speed, and sample data is as shown in table 1, classification data " -1 " representing fault is pre-
Alert point, " 1 " represent normal point
1 sample data of table
The process that data set classification work is carried out using SVM is with previously selected some Nonlinear Mappings first by the input space
Be mapped to high-dimensional feature space, cutter head torque and fltting speed in two-dimensional space be it is discrete, belong to linearly inseparable, such as Fig. 1
It is shown.X-axis represents fltting speed, and Y-axis represents cutter head torque, G1 representing fault point, and G2 represents normal point.
As shown in Fig. 2, creating a band using gaussian kernel function, there are two the SVM support vector machines of input variable.Such as Fig. 3
The parameter of shown adjustment gaussian kernel function carries out SVM support vector machines and sample input/output argument using SMO learning algorithm
Training, obtains categorised decision function i.e. final SVM model.Fig. 4 is the supporting vector after training.Fig. 5 is using final
SVM model original sample is tested, G1 Hit represents the fault point that recognizes, and G2 Hit representative recognizes normal
Point, G1 Miss represent the unidentified fault point arrived, and G2 Miss represents the unidentified normal point arrived.
(3) TBM fault early warning method
The real time data of shield TBM fault pre-alarming process shown in fig. 6, shield TBM key parameter is screened by fault model, is protected
The data of shield TBM key parameter needed for staying model.
The fault model of Fig. 1 is made of cutter head torque and fltting speed, i.e., from the shield TBM key parameter number of all acquisitions
Only retain the data of cutter head torque and fltting speed in.Input data of this group of data as the SVM model after training, passes through
SVM category of model identification after training, the type (normal point/fault point) of output reorganization data, if being identified as normal point
Next group of real time data is then extracted, record and cumulative time if being identified as fault point.Cumulative time is thought more than threshold value
Failure, push warning message to related management personnel.
The technical personnel in the technical field can readily realize the present invention with the above specific embodiments,.But it answers
Work as understanding, the present invention is not limited to above-mentioned specific embodiments.On the basis of the disclosed embodiments, the technical field
Technical staff can arbitrarily combine different technical features, to realize different technical solutions.
Claims (5)
1. a kind of shield TBM fault early warning method based on SVM algorithm, it is characterised in that: specific step is as follows for this method:
(1) shield TBM key boring parameter is created according to previous generation problem in shield TBM work progress
Shield TBM fault model；
(2) shield TBM machine learning classification algorithm SVM modeled by analysis selection shield-tunneling construction data sample, trained
It is verified with classification；
(3) Classification and Identification is carried out to shield TBM construction real time data using the SVM model after training, and fault message is carried out
Determine and pushes.
2. the shield TBM fault early warning method according to claim 1 based on SVM algorithm, it is characterised in that: the step
(1) the shield TBM key boring parameter in refers to the shields such as cutterhead revolving speed, cutter head torque, gross thrust, fltting speed, soil pressure TBM
Two or more combination in crucial boring parameter.
3. the shield TBM fault early warning method according to claim 1 based on SVM algorithm, it is characterised in that: the step
(2) detailed process is: it is defeated as the input of SVM that the shield TBM fault model created according to step (1) prepares sample data
Then parameter out adjusts kernel function first according to one SVM support vector machines of number syncaryon function creation of input variable
Parameter is trained SVM support vector machines and sample input/output argument using SMO learning algorithm, obtains categorised decision letter
Number i.e. final SVM model finally carry out test verifying to sample is prepared using final SVM model.
4. the shield TBM fault early warning method according to claim 1 based on SVM algorithm, it is characterised in that: the step
(3) detailed process is: using step (3) obtained test verifying after SVM model to shield TBM construct real time data into
Row Classification and Identification stores record and cumulative time if being identified as fault point, if the cumulative time is more than threshold value, then it is assumed that
Then failure pushes fault pre-alarming information to related management personnel.
5. the shield TBM fault early warning method according to claim 3 based on SVM algorithm, it is characterised in that: the core
Function is linear kernel function, Polynomial kernel function, gaussian kernel function, Laplce's kernel function, any in sigmod kernel function
It is a kind of.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811589422.1A CN109726463A (en) | 2018-12-25 | 2018-12-25 | A kind of shield TBM fault early warning method based on SVM algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811589422.1A CN109726463A (en) | 2018-12-25 | 2018-12-25 | A kind of shield TBM fault early warning method based on SVM algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109726463A true CN109726463A (en) | 2019-05-07 |
Family
ID=66296310
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811589422.1A Pending CN109726463A (en) | 2018-12-25 | 2018-12-25 | A kind of shield TBM fault early warning method based on SVM algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109726463A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112131530A (en) * | 2020-09-24 | 2020-12-25 | 盾构及掘进技术国家重点实验室 | Big data-based calculation and early warning method for slag discharge amount of earth pressure shield |
| CN114707669A (en) * | 2022-06-02 | 2022-07-05 | 湖南师范大学 | Hob fault diagnosis model training method, diagnosis device and electronic equipment |
| CN116029555A (en) * | 2023-03-22 | 2023-04-28 | 西南科技大学 | Bridge risk identification early warning system based on lightweight neural network and application method |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104102773A (en) * | 2014-07-05 | 2014-10-15 | 山东鲁能软件技术有限公司 | Equipment fault warning and state monitoring method |
| CN104457973A (en) * | 2014-12-29 | 2015-03-25 | 中铁隧道集团有限公司 | Vibration signal detection and fault diagnosis system in shield tunneling process |
| CN106401597A (en) * | 2016-10-27 | 2017-02-15 | 华中科技大学 | Failure prediction and diagnosis control method applicable to shield tunneling machine |
| CN106682781A (en) * | 2016-12-30 | 2017-05-17 | 山东鲁能软件技术有限公司 | Power equipment multi-index prediction method |
| CN107503757A (en) * | 2017-09-20 | 2017-12-22 | 中建交通建设集团有限公司 | A kind of shield tunnel intelligence construction accessory system and application method based on big data technology |
| CN107634857A (en) * | 2017-09-22 | 2018-01-26 | 安徽师范大学 | Fault Model structure and appraisal procedure based on SVM |
-
2018
- 2018-12-25 CN CN201811589422.1A patent/CN109726463A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104102773A (en) * | 2014-07-05 | 2014-10-15 | 山东鲁能软件技术有限公司 | Equipment fault warning and state monitoring method |
| CN104457973A (en) * | 2014-12-29 | 2015-03-25 | 中铁隧道集团有限公司 | Vibration signal detection and fault diagnosis system in shield tunneling process |
| CN106401597A (en) * | 2016-10-27 | 2017-02-15 | 华中科技大学 | Failure prediction and diagnosis control method applicable to shield tunneling machine |
| CN106682781A (en) * | 2016-12-30 | 2017-05-17 | 山东鲁能软件技术有限公司 | Power equipment multi-index prediction method |
| CN107503757A (en) * | 2017-09-20 | 2017-12-22 | 中建交通建设集团有限公司 | A kind of shield tunnel intelligence construction accessory system and application method based on big data technology |
| CN107634857A (en) * | 2017-09-22 | 2018-01-26 | 安徽师范大学 | Fault Model structure and appraisal procedure based on SVM |
Non-Patent Citations (1)
| Title |
|---|
| 刘金刚: "基于PCA 和SVM 的盾构液压系统故障诊断", 计算机仿真 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112131530A (en) * | 2020-09-24 | 2020-12-25 | 盾构及掘进技术国家重点实验室 | Big data-based calculation and early warning method for slag discharge amount of earth pressure shield |
| CN112131530B (en) * | 2020-09-24 | 2024-03-19 | 盾构及掘进技术国家重点实验室 | Calculation and early warning method for slag discharge amount of earth pressure shield based on big data |
| CN114707669A (en) * | 2022-06-02 | 2022-07-05 | 湖南师范大学 | Hob fault diagnosis model training method, diagnosis device and electronic equipment |
| CN116029555A (en) * | 2023-03-22 | 2023-04-28 | 西南科技大学 | Bridge risk identification early warning system based on lightweight neural network and application method |
| CN116029555B (en) * | 2023-03-22 | 2023-06-13 | 西南科技大学 | Bridge risk identification early warning system based on lightweight neural network and application method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhou et al. | Data based complex network modeling and analysis of shield tunneling performance in metro construction | |
| CN109726463A (en) | A kind of shield TBM fault early warning method based on SVM algorithm | |
| CN106778010B (en) | TBM cutter life prediction method based on data-driven support vector regression machine | |
| CN102918423B (en) | Method for earthquake hydrocarbon system anlysis | |
| CN102981182B (en) | 2D seismic data all-horizon automatic tracking method based on unsupervised classification | |
| CN111160571B (en) | Method for driving inversion of geology based on earth pressure balance shield machine parameter data | |
| CN108763752A (en) | The method that water channel determines construction boring parameter is worn under a kind of shield tunnel | |
| CN104462184A (en) | Large-scale data abnormity recognition method based on bidirectional sampling combination | |
| CN104850732A (en) | Oil reservoir small layer partitioning method and device based on sand body statistics | |
| CN107633325A (en) | A kind of shale gas pressure break down-hole accident early diagnosis alarm method and device | |
| CN112100727A (en) | Early warning prevention and control method for water inrush of water-rich tunnel under influence of fault fracture zone | |
| CN111738371B (en) | Stratum fracture number prediction method based on random forest data mining | |
| CN105321204A (en) | Three-dimensional geological raster body model construction method | |
| CN110163965A (en) | A kind of solid threedimensional model building method of monoclinal strate tectonic landform | |
| CN109507738A (en) | Using the more section joint interpretation methods of ground penetrating radar detection underground disease | |
| CN109635879A (en) | A kind of Malfunction Diagnosis for Coal-Mining Machine system that parameter is optimal | |
| CN110197013B (en) | Riverbed bedrock surface modeling method based on Morphing | |
| CN105425291B (en) | A kind of Dominant Facies method that ancient geoaraply recovers | |
| CN113656868A (en) | BIM technology-based hospital construction collaborative management platform | |
| CN101187914A (en) | Damage-free detection archeological method | |
| Lin et al. | A novel domain adversarial time-varying conditions intervened neural network for drill bit wear monitoring of the jumbo drill under variable working conditions | |
| CN106777068A (en) | A kind of farmland quality grade database increment updating method | |
| CN110119522B (en) | Stability rating method for excavated rock slope damage risk analysis | |
| CN107122895A (en) | A kind of method that can realize different rock mass classifications simultaneously | |
| Tian et al. | Application of multidimensional association mining in shield tunneling optimization |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190507 |