CN110348639A - A kind of coal mine roof plate gushing water danger classes prediction technique - Google Patents
A kind of coal mine roof plate gushing water danger classes prediction technique Download PDFInfo
- Publication number
- CN110348639A CN110348639A CN201910638323.6A CN201910638323A CN110348639A CN 110348639 A CN110348639 A CN 110348639A CN 201910638323 A CN201910638323 A CN 201910638323A CN 110348639 A CN110348639 A CN 110348639A
- Authority
- CN
- China
- Prior art keywords
- data
- neuron
- input
- hidden
- 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.)
- Pending
Links
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 title claims abstract description 41
- 239000003245 coal Substances 0.000 title claims abstract description 37
- 238000000034 method Methods 0.000 title claims abstract description 35
- 238000013528 artificial neural network Methods 0.000 claims abstract description 32
- 238000012549 training Methods 0.000 claims abstract description 26
- 238000000513 principal component analysis Methods 0.000 claims abstract description 13
- 230000005540 biological transmission Effects 0.000 claims abstract description 4
- 238000003062 neural network model Methods 0.000 claims abstract description 4
- 238000010606 normalization Methods 0.000 claims abstract description 3
- 210000002569 neuron Anatomy 0.000 claims description 48
- 239000011159 matrix material Substances 0.000 claims description 16
- 210000005036 nerve Anatomy 0.000 claims description 5
- 230000009467 reduction Effects 0.000 claims description 5
- 238000013138 pruning Methods 0.000 claims description 4
- 238000009825 accumulation Methods 0.000 claims description 3
- 230000006978 adaptation Effects 0.000 claims description 2
- 238000004422 calculation algorithm Methods 0.000 claims description 2
- 238000004364 calculation method Methods 0.000 claims description 2
- 230000006870 function Effects 0.000 claims description 2
- 239000004615 ingredient Substances 0.000 claims description 2
- 210000002364 input neuron Anatomy 0.000 claims description 2
- 210000004218 nerve net Anatomy 0.000 claims description 2
- 238000013139 quantization Methods 0.000 abstract description 2
- 230000001737 promoting effect Effects 0.000 abstract 1
- 230000006872 improvement Effects 0.000 description 2
- 230000001537 neural effect Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000002490 cerebral effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000002405 diagnostic procedure Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 239000011435 rock Substances 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000000946 synaptic effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- 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
- 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 relates to technical field of mine safety, in particular to a kind of coal mine roof plate gushing water danger classes prediction technique.The method is based on PCA-LVQ neural network, specifically includes the following steps: step 1: the distance and corresponding danger classes of collection coalfield effective thickness, specific capacity, water-bearing layer load transmission coefficient, effective protection thickness, fractal dimension and key stratum to coal seam；Step 2: principal component analysis and normalization are carried out to data；Step 3: learning vector quantization Neural Network Training Parameter is configured；Step 4: hidden layer network node is chosen；Step 5: establishing neural network model using software；Step 6: the neural network of foundation is trained；Step 7: neural network accuracy of the location estimate.Prediction result precision of the present invention is high, and structure is simple, can effectively prevent the generation of coal mine roof plate water inrush accident, is suitable for promoting and applying.
Description
Technical field
The present invention relates to technical field of mine safety, in particular to a kind of coal mine roof plate gushing water danger classes prediction technique.
Background technique
With the continuous improvement of coal resources in China exploitation rate, exploits difficulty and continue to increase, mining environment is increasingly disliked
Bad, mine safety accident especially coal mine roof plate pressure rack, water inrush accident occur repeatedly, thus cause a large amount of casualties with
And property loss.According to statistics, between nearly 30 years, more than 1600 coal mine water inrush accidents are had occurred in China altogether, and in addition there are also a large amount of mines
It is faced with the threat of roof gushing water.Coal mine roof plate gushing water is that water producing fractures existing for roof overlying rock are communicated to
When artesian aquifer, water body is flowed into drive by crack, causes water inrush accident.For the generation for preventing water inrush accident, containing
When being exploited under water layer, need to reserve certain thickness waterproof coal petrography to ensure safety.But as coal seam upper extraction limit is continuous
It improves, water inrush accident constantly occurs.
The generation of coal mine roof plate water inrush accident in order to prevent needs to design a kind of coal mine roof plate gushing water danger classes pre- now
Survey method.Neural network is that a kind of application is similar to the structure of cerebral nerve Synaptic junction and carries out the mathematical model of information processing,
It is with very strong nonlinear fitting ability.Learning rules are simple, are easy to implement, thus are widely used.But it is to input number
According to there is higher requirement, if input vector is too many, data redundancy will cause, network structure is excessively complicated, influences its instruction instead
Practice precision.
Summary of the invention
To solve the shortcomings of the prior art, the present invention provides a kind of coal mine roof plate gushing water danger classes prediction technique,
Specifically a kind of coal mine roof plate gushing water danger classes prediction technique based on PCA-LVQ neural network.
The technical solution of the present invention is as follows:
A kind of coal mine roof plate gushing water danger classes prediction technique, the method are based on PCA-LVQ neural network, specifically include
Following steps:
Step 1: collecting coalfield related data, including effective thickness, water-bearing layer load transmission coefficient, specific capacity, have
Effect protects thickness, fractal dimension and main key stratum to 6 influence factors of distance in coal seam and corresponding gushing water danger classes.
Step 2: the data of collection being handled, including principal component analysis (PCA) and normalization, utilize principal component point
Analysis calculates the contribution margin of 6 influence factors and accumulation contribution rate is taken to be greater than 85% data, reduces redundancy, realizes dimensionality reduction.Meanwhile
Data after dimensionality reduction are normalized so that because caused by data unit is different error it is minimum, accelerate network convergence and speed.
Step 3: network training parameter setting, occurrence are as follows: error of fitting target value is 0.01, frequency of training 200.
Step 4: intermediate node is chosen, and filters out hidden layer neurode value range using formula, then utilizes beta pruning method
The training data mean square error and prediction data accuracy of different hidden nodes are calculated, finally determines number of nodes.
Step 5: neural network, the related data determined by the dimension and step 3 of inputoutput data and 4,
Neural network model is set up using MATLAB software.
Step 6: the neural network set up using training data training is found optimal weight and threshold value, established
Optimal neural network.
Step 7: utilizing trained neural network, prediction data is predicted, its precision of prediction is examined.
Further, the specific method is as follows for step 2:
Step is 1.: constructing normalized matrix X using initial data；
In formula, n representative sample number, xijJ-th of parameter of as i-th of sample
Step is 2.: the correlation matrix of normalized matrix
Related coefficient calculation formula is as follows:
In formula: rkjFor variable xkAnd xjBetween related coefficient,It is the average value of the i variable,It is j-th of variable
Average value, n is sample size；
Step is 3.: the characteristic equation for calculating correlation matrix R obtains characteristic value and feature vector；
Step is 3.: calculating its contribution rate τ using following formulaiWith contribution rate of accumulative total η:
In formula: λ is matrix exgenvalue
Wherein, principal component number is chosen by calculating the contribution rate of accumulative total of principal component, wherein contribution rate is for indicating comprehensive
Close the ability that variable explains original variable；Contribution rate of accumulative total is bigger to illustrate that data information loss is smaller；
Step is 3.: data band being become owner of and calculates principal component specific value in ingredient expression formula, obtains new input data；
Step is 6.: the data after principal component analysis being normalized, make its data in (0,1) range.
Further, the specific method is as follows for step 4:
Step is 1.: by inputoutput data dimension, determining input layer number and output layer number of nodes；
Step is 2.: utilizing empirical equationHidden neuron node value range is determined,
Wherein, n --- input layer number；M --- output layer number of nodes, a --- 1 to 10 arbitrary numbers；
Step is 3.: using beta pruning method, calculates training data mean square error and the forecast set error of different hidden layers successively to select
Fixed optimal the number of hidden nodes；
Further, the specific method is as follows for step 5:
Step is 1.: netinit initializes connection weight ω and learning rate between each layer；
Step is 2.: setting up neural network using MATLAB software, target type is constructed adaptation nerve using function
The output matrix of network output mode, network error of fitting target value are 0.01, and frequency of training 200, learning rate is using silent
Recognize value 0.01.
Further, the specific method is as follows for step 6:
Step is 1.: the input data after principal component analysis is input to input layer；
Step is 2.: calculating hidden layer weight and input vector X=[x according to the following formula1,x2,x3,…,xQ]TBetween away from
From:
In formula, wijThe weight between input neuron i and intrerneuron j；Q is input vector dimension；
Step is 3.: selection and input vector are apart from the smallest two hidden neurons i, j；
Step is 4.: calculate whether neuron i and neuron j meets following two condition:
1. the corresponding Type-Inconsistencies of hidden neuron i and j；
2. the distance d of hidden neuron i and j and current input modeiAnd djMeet following formula:
P: the window width for the midplane that input vector may be dropped into, usually 2/3 or so, then:
1. if the corresponding classification C of neuron iiClassification C corresponding with input vectorxUnanimously, i.e. Ci=Cx, then neuron i and
The weight of neuron j is modified as follows:
wi_new=wi_old+α(x-wi_old)
In formula, α is learning efficiency；wi_oldFor the weight for adjusting preceding connection neuron i and neuron j；
2. if the corresponding classification C of neuron jjClassification C corresponding with input vectorxUnanimously, i.e. Cj=Cx, then neuron i and
The weight of neuron j is modified as follows:
wi_new=wi_old-α(x-wi_old)
Step is 5.: if hidden neuron i and j are unable to reach requirement, the connection weight of hidden layer related Neurons is updated,
Update mode is as the step 4 in LVQ algorithm.
Further, its weight of step 6 and adjusting thresholds are as follows:
Wherein, p is the input pattern of R dimension, S1It is 4, IW for hidden neuron number1,1Between input layer and hidden layer
Connection weight matrix, n1For the input of hidden neuron, a1For the output of hidden neuron, LW2,1Between hidden layer and output layer
Connection weight matrix, n2For the input of output layer neuron, a2For the output of output layer neuron.
Further, in step 7: the forecast sample of identical dimensional is input to the input layer of trained neural network,
Output result after neural network computing is compared with actual result, calculates its accuracy.
The beneficial effects obtained by the present invention are as follows are as follows:
The present invention is based on the coal mine roof plate gushing water danger classes prediction techniques of PCA-LVQ neural network, can effectively subtract
The excessive problem of few input vector, so that network structure is simpler, accuracy is higher, is conducive to improve coal mine by this method
The accuracy and speed of roof water inrush danger classes prediction, structure is simple, can effectively prevent the hair of coal mine roof plate water inrush accident
It is raw.
Detailed description of the invention
Fig. 1 is LVQ network mode figure；
Fig. 2 (a) is training data mean square error figure；
Fig. 2 (b) is prediction data accuracy figure；
Fig. 3 (a) LVQ neural metwork training result figure；
Fig. 3 (b) PCA-LVQ neural metwork training result figure；
The analysis of Fig. 4 (a) Fisher prediction result；
Fig. 4 (b) LVQ neural network prediction；
Fig. 4 (c) PCA-LVQ neural network prediction.
Specific embodiment
For convenient for it will be understood by those skilled in the art that the present invention, illustrates specific embodiment party of the invention with reference to the accompanying drawing
Formula.
A kind of coal mine roof plate gushing water danger classes prediction technique based on PCA-LVQ neural network of the invention.Coal mine top
One of an important factor for plate gushing water is threat safety of coal mines.In order to quickly and accurately identify coal mine roof plate gushing water danger classes, mention
A kind of coal mine roof plate gushing water danger classes prediction technique based on PCA-LVQ neural network is gone out.By consulting relevant data,
Find out relevant data, including effective thickness, specific capacity, water-bearing layer load transmission coefficient, effective protection thickness, FRACTAL DIMENSION
Value and main key stratum carry out dimensionality reduction to data by principal component analysis (PCA), and use learning vector quantization to the distance in coal seam
(LVQ) neural network predicts its danger classes.Simulation results show proposition based on PCA-LVQ neural network
Coal mine roof plate gushing water danger classes prediction technique to be compared to other diagnostic method accuracys higher.
The present invention has collected 52 groups of coal mine data altogether, and every group of data include 6 groups of variables, therefrom extracts 10 groups of data and carries out in advance
It surveys, remainder data is trained network.
The PCA-LVQ neural network model of proposition and other two kinds of prediction models (Fisher and LVQ) results are carried out pair
Than comparison result is as shown in table 1.
1 prediction result of table compares
It is higher and can reach to can be seen that the more existing two methods accuracy rate of method of the invention from the data in table 1
100% accuracy rate.
Table 2 is the characteristic value and contribution rate after principal component analysis.
2 characteristic value of table and its contribution rate
As shown in table 2, the present invention chooses the part that accumulation characteristic value is greater than 85%, therefore chooses first four as network
Input parameter.
Fig. 1 is neural network structure figure, and as shown in Figure 1, this neural network is by three layers of input layer, hidden layer and output layer mind
It is formed through member.Neuron between input layer and hidden layer is connected with each other, according to training data prediction in network training process
The weight connected between two layers is modified in accuracy.One group of connection of each output layer neuron and hidden neuron, connection
Weight is fixed value 1, and training mode is when input is sent to network, and the hidden neuron nearest from input pattern is swashed
Hair, i.e., excited neuron, playing output valve is 1, and other hidden neuron output valves are 0.It is connected with group where excitor nerve member
The output layer neuron output valve connect is 1, and other output layer neuron output valves are " 0 ", therefore can be identified current defeated
Enter the classification of mode, to judge it.
Fig. 2 (a) is respectively (b) that the corresponding training data mean square error of different the number of hidden nodes and prediction data are accurate
Degree, by Fig. 2 (a), (b) as can be seen that its network is at one's best when the number of hidden nodes is 10, therefore its hidden node is selected as
10。
Fig. 3 (a) is respectively (b) LVQ and PCA-LVQ network training as a result, by Fig. 3 (a), (b) it is found that LVQ nerve net
Permissible accuracy is still not up to after 200 step of network training, error minimum value is 0.015873 when training 43 step, and is carried out
LVQ network training after principal component analysis has reached permissible accuracy when 164 step.
Fig. 4 (a) is respectively (c) (b) that Fisher analysis and LVQ and PCA-LVQ neural network forecast result are obtained with actual value
Compare, wherein ordinate 1 represents safety zone, and 2 represent moderate risk area, and 3 represent danger area.
The embodiments of the present invention described above are not intended to limit the scope of the present invention.It is any in the present invention
Spirit and principle within made modifications, equivalent substitutions and improvements etc., should be included in claim protection model of the invention
Within enclosing.
Claims (7)
1. a kind of coal mine roof plate gushing water danger classes prediction technique, which is characterized in that the method is based on PCA-LVQ nerve net
Network, specifically includes the following steps:
Step 1: collecting coalfield related data, including effective thickness, water-bearing layer load transmission coefficient, specific capacity, effectively guarantor
Thickness, fractal dimension and main key stratum are protected to 6 influence factors of distance in coal seam and corresponding gushing water danger classes；
Step 2: the data of collection being handled, including principal component analysis and normalization, calculate 6 using principal component analysis
The contribution margin of influence factor simultaneously takes accumulation contribution rate to be greater than 85% data, reduces redundancy, realizes dimensionality reduction；Meanwhile to dimensionality reduction after
Data be normalized so that because caused by data unit is different error it is minimum, accelerate network convergence and speed；
Step 3: network training parameter setting, occurrence are as follows: error of fitting target value is 0.01, frequency of training 200；
Step 4: hidden node is chosen, and filters out hidden layer neurode value range using formula, is then calculated using beta pruning method
The training data mean square error and prediction data accuracy of different hidden nodes out finally determines number of nodes；
Step 5: neural network, the related data determined by the dimension and step 3 of inputoutput data and 4 utilize
MATLAB software sets up neural network model；
Step 6: the neural network set up using training data training is found optimal weight and threshold value, established optimal
Neural network；
Step 7: utilizing trained neural network, prediction data is predicted, its precision of prediction is examined.
2. a kind of coal mine roof plate gushing water danger classes prediction technique according to claim 1, which is characterized in that the step
2 the following steps are included:
Step is 1.: constructing normalized matrix X using initial data；
In formula, n representative sample number, xijJ-th of parameter of as i-th of sample
Step is 2.: the correlation matrix of normalized matrix
Related coefficient calculation formula is as follows:
In formula: rkjFor variable xkAnd xjBetween related coefficient,It is the average value of the i variable,It is the flat of j-th of variable
Mean value, n are sample size；
Step is 3.: the characteristic equation for calculating correlation matrix R obtains characteristic value and feature vector；
Step is 4.: calculating its contribution rate τ using following formulaiWith contribution rate of accumulative total η:
In formula: λ is matrix exgenvalue
Wherein, principal component number is chosen by calculating the contribution rate of accumulative total of principal component, wherein contribution rate is for indicating comprehensive change
Amount explains the ability of original variable；Contribution rate of accumulative total is bigger to illustrate that data information loss is smaller；
Step is 5.: data band being become owner of and calculates principal component specific value in ingredient expression formula, obtains new input data；
Step is 6.: the data after principal component analysis being normalized, make its data in (0,1) range.
3. a kind of coal mine roof plate gushing water danger classes prediction technique according to claim 1, which is characterized in that the step
4 include the following steps:
Step is 1.: by inputoutput data dimension, determining input layer number and output layer number of nodes；
Step is 2.: utilizing empirical equationHidden neuron node value range is determined；Wherein,
N is input layer number；M is output layer number of nodes, and a is 1 to 10 arbitrary numbers；
Step is 3.: using beta pruning method, calculates training data mean square error and the forecast set error of different hidden layers successively to select most
Excellent the number of hidden nodes.
4. a kind of coal mine roof plate gushing water danger classes prediction technique according to claim 1, which is characterized in that the step
5 specifically comprise the following steps:
Step is 1.: netinit initializes connection weight ω and learning rate between each layer；
Step is 2.: setting up neural network using MATLAB software, target type is constructed adaptation neural network using function
The output matrix of output mode, network error of fitting target value are 0.01, and frequency of training 200, learning rate uses default value
0.01。
5. a kind of coal mine roof plate gushing water danger classes prediction technique according to claim 1, which is characterized in that the step
6 include the following steps:
Step is 1.: the input data after principal component analysis is input to input layer；
Step is 2.: calculating hidden layer weight and input vector X=[x according to the following formula1,x2,x3,…,xQ]TThe distance between:
In formula, wijThe weight between input neuron i and intrerneuron j；Q is input vector dimension；
Step is 3.: selection and input vector are apart from the smallest two hidden neurons i, j；
Step is 4.: calculate whether neuron i and neuron j meets following two condition:
1. the corresponding Type-Inconsistencies of hidden neuron i and j；
2. the distance d of hidden neuron i and j and current input modeiAnd djMeet following formula:
P: the window width for the midplane that input vector may be dropped into, usually 2/3 or so, then:
1. if the corresponding classification C of neuron iiClassification C corresponding with input vectorxUnanimously, i.e. Ci=Cx, then neuron i and nerve
The weight of first j is modified as follows:
wi_new=wi_old+α(x-wi_old)
In formula, α is learning efficiency；wi_oldFor the weight for adjusting preceding connection neuron i and neuron j；
2. if the corresponding classification C of neuron jjClassification C corresponding with input vectorxUnanimously, i.e. Cj=Cx, then neuron i and nerve
The weight of first j is modified as follows:
wi_new=wi_old-α(x-wi_old)
Step is 5.: if hidden neuron i and j are unable to reach requirement, updating the connection weight of hidden layer related Neurons, updates
Mode is as the step 4 in LVQ algorithm.
6. a kind of coal mine roof plate gushing water danger classes prediction technique according to claim 1, which is characterized in that the step
6 its weight and adjusting thresholds are as follows:
Wherein, p is the input pattern of R dimension, S1It is 4, IW for hidden neuron number1,1Connection between input layer and hidden layer
Weight matrix, n1For the input of hidden neuron, a1For the output of hidden neuron, LW2,1Company between hidden layer and output layer
Meet weight matrix, n2For the input of output layer neuron, a2For the output of output layer neuron.
7. a kind of coal mine roof plate gushing water danger classes prediction technique according to claim 1, which is characterized in that the step
In 7, the forecast sample of identical dimensional is input to the input layer of trained neural network, it is defeated after neural network computing
Result is compared with actual result out, calculates its accuracy.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910638323.6A CN110348639A (en) | 2019-07-16 | 2019-07-16 | A kind of coal mine roof plate gushing water danger classes prediction technique |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910638323.6A CN110348639A (en) | 2019-07-16 | 2019-07-16 | A kind of coal mine roof plate gushing water danger classes prediction technique |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110348639A true CN110348639A (en) | 2019-10-18 |
Family
ID=68176414
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910638323.6A Pending CN110348639A (en) | 2019-07-16 | 2019-07-16 | A kind of coal mine roof plate gushing water danger classes prediction technique |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110348639A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113742995A (en) * | 2021-07-28 | 2021-12-03 | 淄博矿业集团有限责任公司 | Mine water inflow prediction method and system based on coal mine big data |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103745093A (en) * | 2013-12-25 | 2014-04-23 | 中国矿业大学 | Principal component analysis-extreme learning machine (PCA-ELM) based coal mine water inrush prediction method |
| CN104318254A (en) * | 2014-11-13 | 2015-01-28 | 中国矿业大学(北京) | Quick coal and rock recognition method based on DCT low-frequency component characteristics |
| CN107729716A (en) * | 2017-11-27 | 2018-02-23 | 西安建筑科技大学 | A kind of mine water inrush Forecasting Methodology based on long Memory Neural Networks in short-term |
| CN107766816A (en) * | 2017-10-18 | 2018-03-06 | 河海大学 | A kind of Mechanical Failure of HV Circuit Breaker recognition methods based on LVQ neutral nets |
-
2019
- 2019-07-16 CN CN201910638323.6A patent/CN110348639A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103745093A (en) * | 2013-12-25 | 2014-04-23 | 中国矿业大学 | Principal component analysis-extreme learning machine (PCA-ELM) based coal mine water inrush prediction method |
| CN104318254A (en) * | 2014-11-13 | 2015-01-28 | 中国矿业大学(北京) | Quick coal and rock recognition method based on DCT low-frequency component characteristics |
| CN107766816A (en) * | 2017-10-18 | 2018-03-06 | 河海大学 | A kind of Mechanical Failure of HV Circuit Breaker recognition methods based on LVQ neutral nets |
| CN107729716A (en) * | 2017-11-27 | 2018-02-23 | 西安建筑科技大学 | A kind of mine water inrush Forecasting Methodology based on long Memory Neural Networks in short-term |
Non-Patent Citations (5)
| Title |
|---|
| ，俞金寿等: "《软测量技术及其在石油化工中的应用》", 30 June 2000, 北京：化学工业出版社 * |
| 卓金武等: "《MATLAB数学建模方法与实践 第3版》", 31 July 2018, 北京航空航天大学出版社 * |
| 李培: "基于PCA-ELM的预测模型在煤矿突水预测中的应用", 《中国优秀硕士学位论文全文数据库工程科技Ⅰ辑》 * |
| 胡红等: "基于主成分分析和学习矢量化的神经网络岩性识别方法", 《测井技术》 * |
| 闻新等: "《应用MATLAB实现神经网络》", 30 June 2015, 北京：国防工业出版社 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113742995A (en) * | 2021-07-28 | 2021-12-03 | 淄博矿业集团有限责任公司 | Mine water inflow prediction method and system based on coal mine big data |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104732070B (en) | A kind of rock burst grade prediction technique based on information vector machine | |
| CN106650767B (en) | Flood forecasting method based on cluster analysis and real-time correction | |
| Fattahi | Prediction of slope stability using adaptive neuro-fuzzy inference system based on clustering methods | |
| CN112949202A (en) | Bayesian network-based rockburst probability prediction method | |
| CN108133225A (en) | A kind of icing flashover fault early warning method based on support vector machines | |
| Wang et al. | Flood control operations based on the theory of variable fuzzy sets | |
| CN108090677B (en) | Reliability evaluation method for key infrastructure | |
| CN110738355A (en) | urban waterlogging prediction method based on neural network | |
| Seckin | Modeling flood discharge at ungauged sites across Turkey using neuro-fuzzy and neural networks | |
| CN103366062B (en) | Based on the core backbone framework construction method of BBO algorithm and Grid survival | |
| CN111105097B (en) | Dam deformation prediction system and method based on convolutional neural network | |
| CN111062464A (en) | Power communication network reliability prediction and guarantee method and system based on deep learning | |
| CN115063020A (en) | Cascade hydropower station multidimensional safety scheduling device and method based on risk monitoring fusion | |
| CN112000015A (en) | Intelligent BIT design method for heavy-duty gas turbine control system controller module based on LSTM and bio-excitation neural network | |
| Li et al. | Risk analysis of dam based on artificial bee colony algorithm with fuzzy c-means clustering | |
| CN107122860A (en) | Bump danger classes Forecasting Methodology based on grid search and extreme learning machine | |
| CN108390380A (en) | A kind of transformer state parameter trend forecasting method and system | |
| CN114154766B (en) | Method and system for early warning vulnerability of power grid under dynamic prediction of thunder and lightning | |
| CN116090819A (en) | Power distribution network risk situation prediction method based on association rule | |
| CN112926786B (en) | Shallow lake target water level reverse prediction method and system based on association rule model and numerical simulation | |
| CN108280289A (en) | Bump danger classes prediction technique based on local weighted C4.5 algorithms | |
| CN110348639A (en) | A kind of coal mine roof plate gushing water danger classes prediction technique | |
| CN114897007A (en) | Drilling machine health condition evaluation method of composite information layered convolution neural network | |
| CN117726939A (en) | Hyperspectral image classification method based on multi-feature fusion | |
| Xu et al. | RETRACTED ARTICLE: Source discrimination of mine water inrush based on Elman neural network globally optimized by genetic algorithm |
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: 20191018 |
|
| RJ01 | Rejection of invention patent application after publication |