CN107122860A - Bump danger classes Forecasting Methodology based on grid search and extreme learning machine - Google Patents
Bump danger classes Forecasting Methodology based on grid search and extreme learning machine Download PDFInfo
- Publication number
- CN107122860A CN107122860A CN201710290857.5A CN201710290857A CN107122860A CN 107122860 A CN107122860 A CN 107122860A CN 201710290857 A CN201710290857 A CN 201710290857A CN 107122860 A CN107122860 A CN 107122860A
- Authority
- CN
- China
- Prior art keywords
- bump
- learning machine
- extreme learning
- influence factor
- danger classes
- 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
- 238000000034 method Methods 0.000 title claims abstract description 51
- 230000004913 activation Effects 0.000 claims abstract description 32
- 210000002569 neuron Anatomy 0.000 claims abstract description 29
- 238000002790 cross-validation Methods 0.000 claims abstract description 11
- 238000012549 training Methods 0.000 claims description 19
- 239000003245 coal Substances 0.000 claims description 13
- 230000006870 function Effects 0.000 claims description 13
- 238000012544 monitoring process Methods 0.000 claims description 8
- 230000008859 change Effects 0.000 claims description 6
- 238000005457 optimization Methods 0.000 claims description 6
- 238000005065 mining Methods 0.000 claims description 4
- 238000013398 bayesian method Methods 0.000 description 4
- 244000287680 Garcinia dulcis Species 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000002068 genetic effect Effects 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000004080 punching Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000005520 cutting process Methods 0.000 description 1
- 238000005553 drilling Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000005284 excitation Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 210000005036 nerve Anatomy 0.000 description 1
- 239000000843 powder Substances 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000000630 rising effect Effects 0.000 description 1
- 208000011580 syndromic disease Diseases 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
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
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/02—Agriculture; Fishing; Forestry; Mining
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- General Physics & Mathematics (AREA)
- Economics (AREA)
- Health & Medical Sciences (AREA)
- Marketing (AREA)
- Life Sciences & Earth Sciences (AREA)
- Tourism & Hospitality (AREA)
- General Health & Medical Sciences (AREA)
- General Business, Economics & Management (AREA)
- Animal Husbandry (AREA)
- Biomedical Technology (AREA)
- Operations Research (AREA)
- Agronomy & Crop Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Marine Sciences & Fisheries (AREA)
- Mining & Mineral Resources (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Primary Health Care (AREA)
- Artificial Intelligence (AREA)
- Quality & Reliability (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention proposes a kind of bump danger classes Forecasting Methodology based on grid search and extreme learning machine, this method：Using the influence factor data of known bump as extreme learning machine input, using bump danger classes as extreme learning machine output, optimize the hidden layer neuron number of extreme learning machine and the type combination of activation primitive using grid data service, corresponding extreme learning machine is set up according to each grid node, the predictablity rate of respective mesh node is determined using ten folding cross-validation methods to each model, selection predictablity rate highest node determines the hidden layer neuron number of extreme learning machine and the type of activation primitive, set up bump danger classes forecast model；By the influence factor data input bump danger classes forecast model of bump to be predicted, bump danger classes predicted value is obtained.This method is simple and easy to do, while ensure that model has good Generalization Capability.
Description
Technical field
The invention belongs to bump Hazard rank electric powder prediction, and in particular to one kind is based on grid search and pole
Limit the bump danger classes Forecasting Methodology of learning machine.
Background technology
Bump is a kind of common dynamic phenomena, with the increase of China's coal-mine mining depth, bump hair
Raw number of times is in rising trend, and its destructiveness is also serious all the more, causes a tremendous loss of lives and property loss, seriously threatens coal
The safety in production of ore deposit, it is therefore necessary to effectively predicted bump danger classes.
The method being predicted is pressed impact to have using the side such as drilling cuttings method, measurement of water-content coefficient of single influence factor at present
Method, but these methods only consider single influence factor, there is the problem of precision of prediction is not high, recently as the hair of artificial intelligence
Exhibition, has many scholars to employ new technology, new method is predicted to bump Hazard rank, wherein there is artificial neural network
Method, GA-ELM methods, Fisher techniquess of discriminant analysis, SVM models etc., the above method achieves lot of research, still
Due to bump complex genesis, and bump data have the features such as non-linear, correlation, it is therefore necessary to continue to explore
New method is predicted to bump Hazard rank.
Extreme learning machine is a kind of Novel learning algorithm of single hidden layer structure, compared to traditional single hidden layer Feedforward Neural Networks
Network, has the advantages that pace of learning is fast, generalization ability is good, regulation parameter is few, and current application is mainly using optimized algorithm to pole
The input weights and hidden layer deviation for limiting learning machine are optimized, if Zhu Zhijie etc. is using genetic algorithm optimization extreme learning machine
Input weights and hidden layer deviation, which are pressed impact, to be predicted, but the performance of extreme learning machine is mainly by hidden layer nerve
The influence of first number and activation primitive is larger, and needs the parameter optimized more when hidden layer neuron number is more；In addition
Fourth China etc. determines excitation function to coal-winning machine using progressive mode using the preferably optimal hidden neuron number of genetic algorithm than choosing
Power is predicted, yet with the type that activation primitive has been secured when carrying out optimizing to hidden layer neuron quantity,
And the weights of input layer and hidden layer are randomly generated with hidden layer threshold value, therefore, it is difficult to ensure the uniqueness of operation result；
In addition the over-fitting problem of model is not taken into full account when the parameter to extreme learning machine is trained process, thus can not be protected
The estimated performance of model of a syndrome.
The content of the invention
In view of the shortcomings of the prior art, the present invention proposes a kind of bump danger based on grid search and extreme learning machine
Dangerous grade prediction technique.
A kind of bump danger classes Forecasting Methodology based on grid search and extreme learning machine, comprises the following steps：
Step 1：Known bump Monitoring Data, the shadow of known bump in acquisition coal mining mine at diverse location
Ring factor data Z=[z1, z2..., zp]T, bump to be predicted influence factor data Z '=[z '1, z
′2..., z 'p]T, and known bump Monitoring Data is classified according to bump earthquake magnitude intensity criteria for classification,
Bump danger classes corresponding with bump influence factor data is obtained, wherein, ziThe bump known to the i-th class
Influence factor data, z 'iFor the influence factor data of the i-th class bump to be predicted, i=1,2 ..., p, p be bump
Influence factor number；
The bump influence factor includes：Coal seam thickness, seam inclination, buried depth, gas density and influence impact ground
The state parameter of pressure；
It is described influence bump state parameter include geological structure situation, seam inclination change, Coal Seam Thickness Change,
Roof control, depressurization phase, sound coal report.
Step 2：Using influence factor data Z=[z of the zscore standardized methods to known bump1,
z2..., zp]TWith influence factor data Z '=[z ' of bump to be predicted1, z '2..., z 'p]TIt is standardized
Processing, the influence factor data X=[x of the known bump after being standardized1, x2..., xp]TAfter standardization
Influence factor data X '=[x ' of bump to be predicted1, x '2... ..., x 'p]T；
Step 3：By the influence factor data X=[x of the known bump after standardization1, x2..., xp]TAnd its
Corresponding bump danger classes is used as training sample set；
Step 4：It regard the influence factor data of the known bump after the standardization that training sample is concentrated as the limit
The input of habit machine, training sample is concentrated corresponding bump danger classes as the output of extreme learning machine, using grid
The hidden layer neuron number of search method optimization extreme learning machine and the type combination of activation primitive, build according to each grid node
Corresponding extreme learning machine is found, the predictablity rate of respective mesh node, choosing is determined using ten folding cross-validation methods to each model
Select predictablity rate highest node and determine the hidden layer neuron number of extreme learning machine and the type of activation primitive, set up punching
Press danger classes forecast model with hitting；
Step 4.1：The interval of grid data service is set, hidden layer neuron is set according to bump influence factor number
Individual number interval, assignment is carried out to the type of activation primitive, set the line number of grid data service as hidden layer neuron number most
Big value, sets the columns of grid data service as the maximum of activation primitive type assignment, sets up search grid；
The integer for being entered as 1~3 of the activation primitive type, be expressed as sigmoid functions, sin functions,
Hardlim functions.
Step 4.2：Using the line number where node as extreme learning machine hidden layer neuron number, by where node
The corresponding activation primitive type of columns as extreme learning machine activation primitive type, after the standardization that training sample is concentrated
Training sample is concentrated corresponding bump to endanger by the influence factor data of known bump as the input of extreme learning machine
Dangerous grade sets up extreme learning machine as the output of extreme learning machine, using ten folding cross-validation methods, calculates present node and sets up
Extreme learning machine predictablity rate；
Step 4.3：Whether judgement currently searches maximum node number, if so, performing step 4.4, otherwise, searches for next
Node, return to step 4.2；
Step 4.4：Choose the maximum model of predictablity rate in the extreme learning machine set up according to all nodes corresponding
Node sets up extreme learning machine as search result according to the corresponding hidden layer neuron number of the node and activation primitive type
Model, obtains bump Hazard rank forecast model；
Step 5：Bump Hazard rank is predicted, by the influence of the bump to be predicted after standardization because
Prime number is according to X '=[x '1, x '2..., x 'p]TBump danger classes forecast model is inputted, bump danger etc. is obtained
Level predicted value.
Beneficial effects of the present invention：
The present invention proposes a kind of bump danger classes Forecasting Methodology based on grid search and extreme learning machine, due to
The complicated mechanism that bump occurs, influence factor is more, and the inventive method is entered using zscore methods to influence factor data
Row standardization eliminates influence of the different dimensions to model；The performance of extreme learning machine is by hidden layer neuron number and activation letter
The influence of several classes of types is larger, and ten folding cross validations are combined to hidden layer neuron number in extreme learning machine using grid data service
And activation primitive type has carried out Combinatorial Optimization, this method is simple and easy to do, while ensure that model has good Generalization Capability.
Brief description of the drawings
Fig. 1 is pre- for the bump danger classes based on grid search and extreme learning machine in the specific embodiment of the invention
The flow chart of survey method.
Embodiment
The specific embodiment of the invention is described in detail below in conjunction with the accompanying drawings.
A kind of bump danger classes Forecasting Methodology based on grid search and extreme learning machine, as shown in figure 1, including
Following steps：
Step 1：Known bump Monitoring Data, the shadow of known bump in acquisition coal mining mine at diverse location
Ring factor data Z=[z1, z2..., zp]T, bump to be predicted influence factor data Z '=[z '1, z
′2..., z 'p]T, and known bump Monitoring Data is classified according to bump earthquake magnitude intensity criteria for classification,
Bump danger classes corresponding with bump influence factor data is obtained, wherein, ziThe bump known to the i-th class
Influence factor data, z 'iFor the influence factor data of the i-th class bump to be predicted, i=1,2 ..., p, p is bump shadow
Ring factor number.
In present embodiment, bump influence factor includes：Coal seam thickness z1, seam inclination z2, buried depth z3, gas it is dense
Spend z4With the state parameter of influence bump.
Influenceing the state parameter of bump includes geological structure situation z5, seam inclination change z6, Coal Seam Thickness Change
z7, roof control z8, depressurization phase z9, ring coal report z10。
Assignment is carried out to the state parameter for influenceing bump：According to the different conditions of the state parameter of influence bump
Its correspondence assignment is set, the state parameter of each influence bump integer value is entered as, as shown in table 1：
Table 1 respectively influences the state parameter assignment of bump
Impact ground pressure monitoring data are classified according to bump earthquake magnitude intensity criteria for classification in present embodiment, obtained
It is 4 grades to bump danger classes, is grade 1 respectively：Microshock, grade 2：Weak impact, grade 3：Medium impact, class 4：
Thump.
In present embodiment, obtain Yanshitai Colliery diverse location at bump Monitoring Data and influence accordingly because
Prime number evidence, the danger classes of bump and corresponding influence factor data as shown in table 2, wherein influence in preceding 26 groups of data because
Prime number presses to respective impact data as known bump danger etc. according to the influence factor data as known bump
Level, influence factor data in rear 10 groups of data as bump to be predicted influence factor data.
The influence factor data of bump at the Yanshitai Colliery diverse location of table 2 and corresponding bump are dangerous
Grade
Step 2：Using influence factor data Z=[z of the zscore standardized methods to known bump1,
z2..., zp]TWith influence factor data Z '=[z ' of bump to be predicted1, z '2..., z 'p]TIt is standardized
Processing, the influence factor data X=[x of the known bump after being standardized1, x2..., xp]TAfter standardization
Influence factor data X '=[x ' of bump to be predicted1, x '2... ..., x 'p]T。
In present embodiment, place is standardized to the influence factor data of bump using zscore standardized methods
Shown in the formula of reason such as formula (1)：
Wherein, xijFor j-th of value of the i-th class bump influence factor data after standardization, μiImpact ground for the i-th class
Press the average of influence factor data, σiFor the standard deviation of the i-th class bump influence factor data, ziiFor the i-th class collected
J-th of value of bump influence factor data, j=1,2 ..., N, N=36 is bump influence factor data count.
In present embodiment, 36 groups of influence factor data in table 2 are standardized, wherein preceding 26 groups of influence factor data
X ' is constituted after composition X after standardization, rear 10 groups of influence factor data normalizations.
Step 3：By the influence factor data X=[x of the known bump after standardization1, x2..., xp]TAnd its
Corresponding bump danger classes is used as training sample set.
Step 4：It regard the influence factor data of the known bump after the standardization that training sample is concentrated as the limit
The input of habit machine, training sample is concentrated corresponding bump danger classes as the output of extreme learning machine, using grid
The hidden layer neuron number of search method optimization extreme learning machine and the type combination of activation primitive, build according to each grid node
Corresponding extreme learning machine is found, the predictablity rate of respective mesh node, choosing is determined using ten folding cross-validation methods to each model
Select predictablity rate highest node and determine the hidden layer neuron number of extreme learning machine and the type of activation primitive, set up punching
Press danger classes forecast model with hitting.
Step 4.1：The interval of grid data service is set, hidden layer neuron is set according to bump influence factor number
Individual number interval, assignment is carried out to the type of activation primitive, set the line number of grid data service as hidden layer neuron number most
Big value, sets the columns of grid data service as the maximum of activation primitive type assignment, sets up search grid.
In present embodiment, setting grid data service at intervals of 1, set implicit according to bump influence factor number
Layer neuron number interval is [1,100], and the type assignment of activation primitive is 1~3 integer, is expressed as sigmoid letters
It is that 1, sin function values are 2, hardlim that sigmoid function values are made in number, sin functions, hardlim functions, the present embodiment
Function value is 3.
The line number that grid data service is set in present embodiment is set to 100 rows, sets the columns of grid data service as 3
Row.
Step 4.2：Using the line number where node as extreme learning machine hidden layer neuron number, by where node
The corresponding activation primitive type of columns as extreme learning machine activation primitive type, after the standardization that training sample is concentrated
Training sample is concentrated corresponding bump to endanger by the influence factor data of known bump as the input of extreme learning machine
Dangerous grade sets up extreme learning machine as the output of extreme learning machine, using ten folding cross-validation methods, calculates present node and sets up
Extreme learning machine predictablity rate.
In present embodiment, using ten folding cross-validation methods, known impact after the standardization that training sample is concentrated
The influence factor data of pressure are divided into ten parts, in turn will wherein 9 parts as training data, 1 part, as test data, is used as the limit
The input of habit machine, by ten computings, calculates ten bump danger classes corresponding with training sample concentration that predict the outcome
Accuracy rate, using the predictablity rate as respective mesh node evaluation index.
Step 4.3：Whether judgement currently searches maximum node number, if so, performing step 4.4, otherwise, searches for next
Node, return to step 4.2.
In present embodiment, maximum node number is 97 rows, 1 row.
Step 4.4：Choose the maximum model of predictablity rate in the extreme learning machine set up according to all nodes corresponding
Node sets up extreme learning machine as search result according to the corresponding hidden layer neuron number of the node and activation primitive type
Model, obtains bump Hazard rank forecast model.
In present embodiment, bump danger classes forecast model is three-decker, as shown in formula (2)：
Wherein, M is hidden layer neuron number, v=1,2 ..., M, ωvFor the connection weight of input layer and hidden layer, βv
For the connection weight of hidden layer and output layer, bvThe extreme learning machine obtained for the threshold value of hidden layer neuron, g (*) for optimization
Activation primitive, okFor the prediction classification of bump danger classes, xkThe training sample of input is concentrated after k-th of standardization
Bump influence factor data, k=1,2 ..., N.
In present embodiment, it is 97, activation primitive to obtain optimal node for 97 rows, 1 row, the i.e. number of hidden layer neuron
Type be sigmoid functions, obtain the part input layer of extreme learning machine model and the weights of hidden layer and hidden layer threshold value b
Such as the institute of table 3 not：
The part input layer of the extreme learning machine model of table 3 and the weights and hidden layer threshold value b of hidden layer
Extreme learning machine model is set up according to node correspondence parameter, by the correct recognition rata of ten folding cross validation models
For 0.84615.
In present embodiment, in order to be compared with institute extracting method, be respectively adopted Nae Bayesianmethod and
AdaboostM1 methods set up bump danger classes forecast model, by ten folding cross validations, the correct recognition rata of model
Respectively 0.7692,0.6154,0.84615 is below, shows that the Predicting Model of Rock Burst set up according to this method has more
Excellent performance.
Step 5：Bump Hazard rank is predicted, by the influence of the bump to be predicted after standardization because
Prime number is according to X '=[x '1, x '2..., x 'p]TBump danger classes forecast model is inputted, bump danger etc. is obtained
Level predicted value.
Using the inventive method, Nae Bayesianmethod and AdaboostM1 methods according to rear 10 groups of influence factors in table 2
Data after data normalization press Hazard rank to be predicted respective impact, predict the outcome as shown in table 4：
Table 4 predicts the outcome
9 groups of bumps in the forecast model Accurate Prediction data set up as seen from the table using context of methods are dangerous
Grade, the medium impact of the 10th group of data is only held up be judged to weak bump, and Nae Bayesianmethod and
Wherein 8 groups grades of the equal Accurate Prediction of AdaboostM1 methods, wherein Nae Bayesianmethod are medium by the 2nd group and the 7th group
Bump is pressed with being mistaken for thump, and AdaboostM1 methods then respectively press the 4th group of microshock, the 7th group medium
Bump is pressed with being mistaken for thump.
Claims (3)
1. a kind of bump danger classes Forecasting Methodology based on grid search and extreme learning machine, it is characterised in that including
Following steps：
Step 1：Obtain known bump Monitoring Data in coal mining mine at diverse location, known bump influence because
Prime number is according to Z=[z1, z2..., zp]T, bump to be predicted influence factor data Z '=[z '1, z '2..., z
′p]T, and known bump Monitoring Data is classified according to bump earthquake magnitude intensity criteria for classification, obtain and impact ground
The corresponding bump danger classes of influence factor data is pressed, wherein, ziThe influence factor number of bump known to the i-th class
According to z 'iFor the influence factor data of the i-th class bump to be predicted, i=1,2 ..., p, p be bump influence factor
Number；
Step 2：Using influence factor data Z=[z of the zscore standardized methods to known bump1, z2..., zp]T
With influence factor data Z '=[z ' of bump to be predicted1, z '2..., z 'p]TIt is standardized, obtains standard
The influence factor data X=[x of known bump after change1, x2..., xp]TWith the impact to be predicted after standardization
Influence factor data X '=[x ' of pressure1, x '2..., x 'p]T；
Step 3：By the influence factor data X=[x of the known bump after standardization1, x2..., xp]TAnd its it is corresponding
Bump danger classes is used as training sample set；
Step 4：It regard the influence factor data of the known bump after the standardization that training sample is concentrated as extreme learning machine
Input, training sample is concentrated to corresponding bump danger classes as the output of extreme learning machine, using grid search
The hidden layer neuron number of method optimization extreme learning machine and the type combination of activation primitive, phase is set up according to each grid node
Extreme learning machine is answered, the predictablity rate of respective mesh node is determined using ten folding cross-validation methods to each model, selection is pre-
Survey accuracy rate highest node and determine the hidden layer neuron number of extreme learning machine and the type of activation primitive, set up impact ground
Press danger classes forecast model；
Step 4.1：The interval of grid data service is set, hidden layer neuron number is set according to bump influence factor number
Interval, the type to activation primitive carries out assignment, sets the line number of grid data service as the maximum of hidden layer neuron number,
The columns of grid data service is set as the maximum of activation primitive type assignment, search grid is set up；
Step 4.2：Using the line number where node as extreme learning machine hidden layer neuron number, by the columns where node
Corresponding activation primitive type as extreme learning machine activation primitive type, known to after the standardization that training sample is concentrated
Training sample is concentrated corresponding bump danger etc. by the influence factor data of bump as the input of extreme learning machine
Level sets up extreme learning machine as the output of extreme learning machine, using ten folding cross-validation methods, calculates the pole that present node is set up
Limit the predictablity rate of learning machine；
Step 4.3：Whether judgement currently searches maximum node number, if so, performing step 4.4, otherwise, searches for next section
Point, return to step 4.2；
Step 4.4：Choose the maximum corresponding node of model of predictablity rate in the extreme learning machine set up according to all nodes
As search result, extreme learning machine mould is set up according to the corresponding hidden layer neuron number of the node and activation primitive type
Type, obtains bump Hazard rank forecast model；
Step 5：Bump Hazard rank is predicted, by the influence factor number of the bump to be predicted after standardization
According to X '=[x '1, x '2..., x 'p]TBump danger classes forecast model is inputted, bump danger classes is obtained pre-
Measured value.
2. the bump danger classes Forecasting Methodology according to claim 1 based on grid search and extreme learning machine,
Characterized in that, the bump influence factor includes：Coal seam thickness, seam inclination, buried depth, gas density and influence impact
The state parameter of ground pressure；
The state parameter of the influence bump includes geological structure situation, seam inclination change, Coal Seam Thickness Change, top plate
Management, depressurization phase, sound coal report.
3. the bump danger classes Forecasting Methodology according to claim 1 based on grid search and extreme learning machine,
Characterized in that, the integer for being entered as 1~3 of the activation primitive type, be expressed as sigmoid functions, sin functions,
Hardlim functions.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710290857.5A CN107122860B (en) | 2017-04-28 | 2017-04-28 | Rock burst danger level prediction method based on grid search and extreme learning machine |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710290857.5A CN107122860B (en) | 2017-04-28 | 2017-04-28 | Rock burst danger level prediction method based on grid search and extreme learning machine |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107122860A true CN107122860A (en) | 2017-09-01 |
| CN107122860B CN107122860B (en) | 2020-01-07 |
Family
ID=59725143
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710290857.5A Expired - Fee Related CN107122860B (en) | 2017-04-28 | 2017-04-28 | Rock burst danger level prediction method based on grid search and extreme learning machine |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107122860B (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110630330A (en) * | 2019-09-23 | 2019-12-31 | 辽宁工程技术大学 | Rock burst classification and judgment method based on energy release main body |
| CN111325461A (en) * | 2020-02-18 | 2020-06-23 | 山东科技大学 | Real-time evaluation method for coal seam impact risk based on vibration monitoring technology |
| CN111764963A (en) * | 2020-07-06 | 2020-10-13 | 中国矿业大学(北京) | Rock burst prediction method based on fast-RCNN |
| CN113009077A (en) * | 2021-02-18 | 2021-06-22 | 南方电网数字电网研究院有限公司 | Gas detection method, gas detection device, electronic apparatus, and storage medium |
| CN113298299A (en) * | 2021-05-13 | 2021-08-24 | 华北科技学院(中国煤矿安全技术培训中心) | BP neural network-based coal bed impact risk intelligent evaluation method |
| CN113469342A (en) * | 2021-07-08 | 2021-10-01 | 北京科技大学 | Rock burst early warning method based on deep learning microseismic monitoring data |
| CN113901939A (en) * | 2021-10-21 | 2022-01-07 | 黑龙江科技大学 | Rock burst danger level prediction method based on fuzzy correction, storage medium and equipment |
| CN114877925A (en) * | 2022-03-31 | 2022-08-09 | 上海交通大学 | Comprehensive energy system sensor fault diagnosis method based on extreme learning machine |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103256073A (en) * | 2013-04-28 | 2013-08-21 | 中国矿业大学 | Underground coal mine impact mine pressure partition grading prediction method |
| CN105785471A (en) * | 2016-02-14 | 2016-07-20 | 辽宁工程技术大学 | Impact danger evaluation method of mine pre-exploiting coal seam |
-
2017
- 2017-04-28 CN CN201710290857.5A patent/CN107122860B/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103256073A (en) * | 2013-04-28 | 2013-08-21 | 中国矿业大学 | Underground coal mine impact mine pressure partition grading prediction method |
| CN105785471A (en) * | 2016-02-14 | 2016-07-20 | 辽宁工程技术大学 | Impact danger evaluation method of mine pre-exploiting coal seam |
Non-Patent Citations (4)
| Title |
|---|
| HUANG GUANGBIN 等: "《Optimizationmethodbasedextremelearningmachineforclassification》", 《ELSEVIER:NEUROCOMPUTING》 * |
| S. BALASUNDARAM • DEEPAK GUPTA: "《On optimization based extreme learning machine in primal for regression and classification by functional iterative method》", 《INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS》 * |
| 朱志洁等: "《基于GA-ELM的冲击地压危险性预测研究》", 《中国安全生产科学技术》 * |
| 李烨等: "《基于改进GS-SVM的煤矿冲击地压预测研究》", 《世界科技研究与发展》 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110630330A (en) * | 2019-09-23 | 2019-12-31 | 辽宁工程技术大学 | Rock burst classification and judgment method based on energy release main body |
| CN111325461A (en) * | 2020-02-18 | 2020-06-23 | 山东科技大学 | Real-time evaluation method for coal seam impact risk based on vibration monitoring technology |
| CN111325461B (en) * | 2020-02-18 | 2022-03-08 | 山东科技大学 | Real-time evaluation method for coal seam impact risk based on vibration monitoring technology |
| CN111764963A (en) * | 2020-07-06 | 2020-10-13 | 中国矿业大学(北京) | Rock burst prediction method based on fast-RCNN |
| CN113009077A (en) * | 2021-02-18 | 2021-06-22 | 南方电网数字电网研究院有限公司 | Gas detection method, gas detection device, electronic apparatus, and storage medium |
| CN113298299A (en) * | 2021-05-13 | 2021-08-24 | 华北科技学院(中国煤矿安全技术培训中心) | BP neural network-based coal bed impact risk intelligent evaluation method |
| CN113469342A (en) * | 2021-07-08 | 2021-10-01 | 北京科技大学 | Rock burst early warning method based on deep learning microseismic monitoring data |
| CN113901939A (en) * | 2021-10-21 | 2022-01-07 | 黑龙江科技大学 | Rock burst danger level prediction method based on fuzzy correction, storage medium and equipment |
| CN113901939B (en) * | 2021-10-21 | 2022-07-01 | 黑龙江科技大学 | Rock burst danger level prediction method based on fuzzy correction, storage medium and equipment |
| CN114877925A (en) * | 2022-03-31 | 2022-08-09 | 上海交通大学 | Comprehensive energy system sensor fault diagnosis method based on extreme learning machine |
| CN114877925B (en) * | 2022-03-31 | 2023-08-22 | 上海交通大学 | Comprehensive energy system sensor fault diagnosis method based on extreme learning machine |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107122860B (en) | 2020-01-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107122860A (en) | Bump danger classes Forecasting Methodology based on grid search and extreme learning machine | |
| CN107122861B (en) | Gas emission quantity prediction method based on PCA-PSO-ELM | |
| CN104794361B (en) | A kind of water-drive pool development effectiveness integrated evaluating method | |
| CN112529341B (en) | Drilling well leakage probability prediction method based on naive Bayesian algorithm | |
| CN103198354B (en) | One utilizes BP neural network and non-dominated sorted genetic algorithm NSGA2 to realize oil-field oil pumper oil recovery energy-saving and production-increase optimization method | |
| CN103617147A (en) | Method for identifying mine water-inrush source | |
| CN103984788B (en) | A kind of coal entry anchor rod support automated intelligent design and optimization system | |
| CN106980877A (en) | A kind of Prediction of Blasting Vibration method based on particle cluster algorithm Support Vector Machines Optimized | |
| CN109934398A (en) | A kind of drill bursting construction tunnel gas danger classes prediction technique and device | |
| CN103226741A (en) | Urban water supply network tube explosion prediction method | |
| CN108229811A (en) | A kind of method for evaluating fracture and vug carbonate reservoir flood effectiveness | |
| CN114757266B (en) | Construction method of rock burst prediction model driven by expert knowledge and data fusion | |
| CN107843920A (en) | A kind of cluster analysis optimizes population reservoir attribute Forecasting Methodology | |
| CN104794550A (en) | WT-KPCA-SVR coupling model based gas emission quantity prediction method | |
| CN104200292A (en) | Forecasting method for height of water-flowing fractured zone | |
| CN103440525A (en) | Urban lake and reservoir water bloom emergency treatment multiple-target multiple-layer decision-making method based on Vague value similarity measurement improved algorithm | |
| CN110489844A (en) | One kind being suitable for the uneven large deformation grade prediction technique of soft rock tunnel | |
| CN118191967B (en) | Intelligent early warning system and method for full-space three-dimensional monitoring of water damage risk of coal seam roof | |
| CN114117881A (en) | Sand production risk prediction method and system | |
| CN115438823A (en) | Borehole wall instability mechanism analysis and prediction method and system | |
| CN115130375A (en) | Rock burst intensity prediction method | |
| CN104680025A (en) | Oil pumping unit parameter optimization method on basis of genetic algorithm extreme learning machine | |
| CN115234220A (en) | Method and device for identifying underground stick-slip vibration in real time by using intelligent drill bit | |
| Yi et al. | Spatial correlation-based machine learning framework for evaluating shale gas production potential: A case study in southern Sichuan Basin, China | |
| CN103628816A (en) | Seismic attribute guidance horizontal-well-steering method and device thereof |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20200107 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |