CN111126658A - Coal mine gas prediction method based on deep learning - Google Patents

Coal mine gas prediction method based on deep learning Download PDF

Info

Publication number
CN111126658A
CN111126658A CN201911121709.6A CN201911121709A CN111126658A CN 111126658 A CN111126658 A CN 111126658A CN 201911121709 A CN201911121709 A CN 201911121709A CN 111126658 A CN111126658 A CN 111126658A
Authority
CN
China
Prior art keywords
data
gas
coal mine
network
deep learning
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
Application number
CN201911121709.6A
Other languages
Chinese (zh)
Inventor
王毅
景毅
张林娟
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Taiyuan University of Technology
Original Assignee
Taiyuan University of Technology
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Taiyuan University of Technology filed Critical Taiyuan University of Technology
Priority to CN201911121709.6A priority Critical patent/CN111126658A/en
Publication of CN111126658A publication Critical patent/CN111126658A/en
Pending legal-status Critical Current

Links

Images

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
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
    • G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • 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
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/084—Backpropagation, e.g. using gradient descent
    • 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—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/02—Agriculture; Fishing; Mining

Abstract

The invention discloses a coal mine gas prediction method based on deep learning, which is characterized in that an observable data set oriented to big data analysis is established according to actual production data collected on site, data preparation of high-dimensional gas data is carried out, the data preparation comprises preprocessing of measure misalignment and missing data, time sequence preprocessing, normalization and dimensionality reduction processing of sample data and the like, a Deep Neural Network (DNN) (deep Neural network) is selected as a gas data perception model, a distributed deep learning framework is built based on Keras, a plurality of machine learning algorithms are integrated, an automatic machine learning engine is further established, a model is trained and a test is completed, and intelligent prediction of coal mine gas is realized by applying the deep Neural network regression model. By the method and the device, more accurate risk pre-judgment basis can be provided for the stoping and tunneling process of the coal mine.

Description

Coal mine gas prediction method based on deep learning
Technical Field
The invention relates to the technical field of coal mine ventilation safety information, in particular to a coal mine gas prediction method based on deep learning.
Background
Coal mine gas prediction is a prerequisite for realizing gas accident prevention, and coal mine gas prediction accuracy directly influences actual application of prediction results. The traditional gas prediction method is limited by data samples, information processing technology, prediction model scale and the like, and the practicability and the accuracy of the traditional gas prediction method cannot meet the actual needs of the site. With the continuous development of mechanization, automation and informatization of coal mine safety production, information related to coal mine underground ventilation safety is gradually accumulated, and a massive high-dimensional information set is formed. However, for historical and technical reasons, the massive data are idle or only subjected to preliminary retrieval analysis, and a large amount of useful information and objective rules implicit behind the massive data are not further mined and discovered. If the big data in the field of coal mine safety production is effectively integrated, the multidimensional data acquired by the monitoring system is subjected to deep correlation analysis and fusion utilization, a more accurate and efficient early warning system is built, open sharing and big data application of mass data resources in the field of safety production are realized, and a more convenient and efficient method is provided for mine safety production.
Disclosure of Invention
The invention solves the technical problems of establishing a knowledge perception model for coal mine production practice and providing a more accurate risk pre-judgment basis for a coal mine recovery and tunneling process by aiming at the key technical problems of a dynamic learning type multi-level trend prediction method.
In order to solve the technical problems, the technical scheme of the invention is as follows: the coal mine gas prediction method based on deep learning is provided, and comprises the following steps:
integrating original ventilation safety information, extracting historical gas characteristic data from a database of mine data characteristics, preprocessing the historical gas characteristic data, and establishing an observable data set of the historical gas characteristic data facing big data analysis;
dividing historical gas characteristic value data in the observable data set into two parts, respectively serving as a training set and a test set, training the constructed deep neural network model DNN through the training set pair, and testing the trained deep neural network model DNN through the test set;
and inputting real-time gas characteristic data into the trained deep neural network model DNN, and obtaining an output result of the deep neural network model DNN as a predicted value of a coal mine gas emission value.
Wherein, the integration of the original ventilation safety information comprises:
screening, analyzing and sorting data aiming at coal mine ventilation safety by using a UML tool, and establishing a data demand model;
and completing data collection, supplement and perfection according to the data demand model.
The method for preprocessing the historical gas characteristic data comprises the following steps:
preprocessing the abnormal mine gas data by a cubic spline interpolation method of particle swarm optimization;
preprocessing gas data time series;
forming and normalizing a gas data sample set;
and performing PCA dimensionality reduction on the high-dimensional gas sequence data.
In the step of preprocessing the gas data time sequence, a sample histogram and an empirical distribution function method are adopted to carry out nonparametric inference on the whole, the distribution rule of the gas concentration whole data is presumed, and meanwhile, a cubic exponential smoothing processing method and an average value correction method are used to respectively process a missing value and an abnormal value of the original time sequence.
When the deep neural network model DNN is trained, an encoding network and a decoding network are constructed, and the encoding network activates a function to train a set sample X0Coded as X ‘ 0(ii) a For a decoding network, coding vectors X 'obtained by coding the network'0Y is predicted by the decoding network so that Y is as identical as possible to the correct tag Y.
The deep neural network model DNN training method comprises the following steps:
establishing an automatic encoder model of a Z layer, defining an activation function of an encoding network as f, outputting the characteristic to be extracted, and performing the encoding process of an automatic encoder of the first layer as follows:
Figure BDA0002275628610000021
Figure BDA0002275628610000022
Figure BDA0002275628610000023
wherein, X0As the input characteristic value, xiThe function is an activation function for the ith input characteristic attribute; w1As a weight matrix between the input layer and the hidden layer, b1Is an offset;
h is to be1As input to the second hidden layer and train W2And b2The same method is carried out until the weight W of the z-th layer is trainedzAnd bz；
Outputting H of the encoded network through the decoding networkzY is retrieved by the decoding network so that the decoded output Y is as equal as possible to that of the correct tag Y.
Wherein, the decoding network utilizes a supervised back propagation algorithm to finely adjust the network parameters of the whole network. The fine tuning process is done by minimizing the reconstruction error E (θ), and the process of parameter update is as follows:
Figure BDA0002275628610000031
Figure BDA0002275628610000032
wherein M is the number of training set samples, y is the output value of the network, TkIs an offset vector, JθThen the cost function is expressed, theta is the singular stack of the layers of the network, theta is { W, b }, and theta is { theta }1,θ2,…,θzThe backpropagation algorithm updates the stacked theta, α is the learning rate.
The invention has the beneficial effects that: according to the coal mine gas emission quantity prediction method based on deep learning, knowledge and rules related to unknown gas emission quantity of human beings are learned by using historical big data through a knowledge perception model based on gas data, the problems that the traditional mine gas emission quantity prediction method is low in precision and cannot meet the requirements of practical application are solved, and medium-short term prediction of the mine gas emission quantity is achieved by using a machine learning method on the basis of the existing big data.
Description of the figures
FIG. 1 is a schematic flow chart of a coal mine gas prediction method based on deep learning according to the present invention;
FIG. 2 is a frame diagram of a coal mine gas intelligent prediction system of a coal mine gas prediction method based on deep learning provided by the invention;
FIG. 3 is a schematic diagram of a prediction result of a deep learning engine of a coal mine gas prediction method based on deep learning provided by the invention.
Detailed Description
The invention is described in further detail below with reference to the accompanying drawings.
Referring to fig. 1, a coal mine gas prediction method based on deep learning is provided, which includes the steps:
integrating original ventilation safety information, extracting historical gas characteristic data from a database of mine data characteristics, preprocessing the historical gas characteristic data, and establishing an observable data set of the historical gas characteristic data facing big data analysis;
dividing historical gas characteristic value data in the observable data set into two parts, respectively serving as a training set and a test set, training the constructed deep neural network model DNN through the training set pair, and testing the trained deep neural network model DNN through the test set;
and inputting real-time gas characteristic data into the trained deep neural network model DNN, and obtaining an output result of the deep neural network model DNN as a predicted value of a coal mine gas emission value.
Wherein, the integration of the original ventilation safety information comprises:
screening, analyzing and sorting data aiming at coal mine ventilation safety by using a UML tool, and establishing a data demand model;
and completing data collection, supplement and perfection according to the data demand model.
The method for preprocessing the historical gas characteristic data comprises the following steps:
preprocessing the abnormal mine gas data by a cubic spline interpolation method of particle swarm optimization;
preprocessing gas data time series;
forming and normalizing a gas data sample set;
and performing PCA dimensionality reduction on the high-dimensional gas sequence data.
In the step of preprocessing the gas data time sequence, a sample histogram and an empirical distribution function method are adopted to carry out nonparametric inference on the whole, the distribution rule of the gas concentration whole data is presumed, and meanwhile, a cubic exponential smoothing processing method and an average value correction method are used to respectively process a missing value and an abnormal value of the original time sequence.
When the deep neural network model DNN is trained, an encoding network and a decoding network are constructed, and the encoding network activates a function to train a set sample X0Coding is X'0(ii) a For a decoding network, coding vectors X 'obtained by coding the network'0Y is predicted by the decoding network so that Y is as identical as possible to the correct tag Y.
The deep neural network model DNN training method comprises the following steps:
establishing an automatic encoder model of a Z layer, defining an activation function of an encoding network as f, outputting the characteristic to be extracted, and performing the encoding process of an automatic encoder of the first layer as follows:
Figure BDA0002275628610000041
Figure BDA0002275628610000042
Figure BDA0002275628610000043
wherein, X0As the input characteristic value, xiThe function is an activation function for the ith input characteristic attribute; w1As a weight matrix between the input layer and the hidden layer, b1Is an offset;
will be provided with1As input to the second hidden layer and train W2And b2The same method is carried out until the weight W of the z-th layer is trainedzAnd bz；
Outputting H of the encoded network through the decoding networkzY is retrieved by the decoding network so that the decoded output Y is as equal as possible to that of the correct tag Y.
Wherein, the decoding network utilizes a supervised back propagation algorithm to finely adjust the network parameters of the whole network. The fine tuning process is done by minimizing the reconstruction error E (θ), and the process of parameter update is as follows:
Figure BDA0002275628610000051
Figure BDA0002275628610000052
wherein M is the number of training set samples, y is the output value of the network, TkIs an offset vector, JθThen the cost function is expressed, theta is the singular stack of the layers of the network, theta is { W, b }, and theta is { theta }1,θ2,…,θzThe backpropagation algorithm updates the stacked theta, α is the learning rate.
According to the embodiment of the invention, main production technical parameters related to gas concentration change in a corresponding time interval are obtained from a coal mine, and through correlation analysis (shown in table 1) among all characteristic parameters, parameters such as the maximum value of upper corner gas concentration, the maximum value of return airway gas quantity, the maximum value of tile drainage roadway gas quantity, the maximum value of outer U roadway gas quantity, gas extraction quantity, daily coal production quantity of a working face, accumulated footage and the like are selected as the characteristics of a neural network to be input; then, after preprocessing of gas abnormal data, preprocessing of a gas data time sequence and dimensionality reduction calculation of high-dimensional gas data, the first 700 groups of samples are selected as training sample sets for training and constructing prediction models in each area, and the rest 220 groups of samples are used as test sample sets for evaluating the performance of the prediction models and analyzing the prediction effect; and finally, respectively extracting the characteristics of the training set and the test set, and inputting the characteristics into the DNN model for model training and testing.
Figure BDA0002275628610000053
Figure BDA0002275628610000061
Table 1 correlation analysis summary table of maximum amount of gas DNN intelligent prediction model training includes: firstly, an automatic encoder model of a Z layer is established, the output of the automatic encoder model can be regarded as a feature to be extracted, and the encoding process of an automatic encoder of a first layer is shown as the following formula:
H1＝f(W1X0+b1)
wherein the function f is an activation function of the coding network, W1,b1For the network weight matrix and offset vector between the first auto-encoder input layer and the hidden layer, θ1＝{W1,b1Is a parameter between the input layer and the hidden layer.
H is to be1As input to the second layer, a network parameter θ via the second layer2To obtain a code vector H2. The same method obtains a code vector Hz. And training layer by using an auto-encoder, taking the output of the Z-layer hidden layer as the input of a decoder, and training the parameters of the decoder by using a supervised method.
Both the encoder and decoder utilize a supervised back propagation algorithm to fine tune the parameters of the neural network. The fine tuning process is done by minimizing the reconstruction error E (θ), and the process of parameter update is as follows:
Figure BDA0002275628610000062
Figure BDA0002275628610000063
wherein M is the number of training set samples, y is the output value of the network, TkIs an offset vector, JθThen the cost function is expressed, theta is the stack of each layer of the network in the singular, theta ═ theta1,θ2,…,θzThe backpropagation algorithm updates the stacked theta, α is the learning rate.
The detailed architecture of the coal mine gas intelligent prediction system based on deep learning is shown in fig. 2 and comprises the following parts: data preprocessing, namely cleaning original data, including processing, normalization and dimension reduction of abnormal data; based on Keras, DNN is selected as a deep neural network model for mine gas prediction, and convergence time is greatly prolonged through super-parameter adjustment; the DNN deep neural network model is characterized in that firstly, a multilayer self-encoder is built for feature extraction, and then a decoder utilizes a back propagation algorithm for parameter fine tuning. And finally, carrying out model test to finish the realization of the whole engine.
Fig. 3 shows the knowledge application effect obtained after the deep learning engine is repeatedly trained, in the figure, the solid line represents the actually measured gas emission amount, the dotted line represents the forecast data of the method, and the actual prediction precision is 85%. As shown in the figure, the forecast result of the engine basically accords with the variation trend of the gas emission quantity, and can reflect the variation amplitude of the emission quantity. The result shows that the local gas emission of a coal field exists regularly, the existing rule can be sensed by a big data machine learning method, and meanwhile, the sensed rule is feasible for forecasting.
Although the present invention has been described with reference to the accompanying drawings, the present invention is not limited to the above embodiments, which are only illustrative and not restrictive, and those skilled in the art can make many modifications without departing from the spirit and scope of the present invention as defined in the appended claims.

Claims (7)

1. A coal mine gas prediction method based on deep learning is characterized by comprising the following steps:
integrating original ventilation safety information, extracting historical gas characteristic data from a database of mine data characteristics, preprocessing the historical gas characteristic data, and establishing an observable data set of the historical gas characteristic data facing big data analysis;
dividing historical gas characteristic value data in the observable data set into two parts, respectively serving as a training set and a test set, training the constructed deep neural network model DNN through the training set pair, and testing the trained deep neural network model DNN through the test set;
and inputting real-time gas characteristic data into the trained deep neural network model DNN, and obtaining an output result of the deep neural network model DNN as a predicted value of a coal mine gas emission value.
2. The deep learning-based coal mine gas prediction method of claim 1, wherein the integrating raw ventilation safety information comprises:
screening, analyzing and sorting data aiming at coal mine ventilation safety by using a UML tool, and establishing a data demand model;
and completing data collection, supplement and perfection according to the data demand model.
3. The deep learning-based coal mine gas prediction method according to claim 1, wherein the step of preprocessing the historical gas characteristic data comprises:
preprocessing the abnormal mine gas data by a cubic spline interpolation method of particle swarm optimization;
preprocessing gas data time series;
forming and normalizing a gas data sample set;
and performing PCA dimensionality reduction on the high-dimensional gas sequence data.
4. The coal mine gas prediction method based on deep learning of claim 3, characterized in that in the step of preprocessing the gas data time series, a sample histogram and an empirical distribution function method are adopted to perform nonparametric inference on the population, so as to estimate the distribution rule of the gas concentration population data, and meanwhile, a cubic exponential smoothing processing method and an average value correction method are used to process the missing value and the abnormal value of the original time series respectively.
5. The coal mine gas prediction method based on deep learning of claim 1, characterized in that, during DNN training, a coding network and a decoding network are constructed, the coding network uses an activation function to train a set sample X0Coding is X'0(ii) a For a decoding network, coding vectors X 'obtained by coding the network'0Y is predicted by the decoding network so that Y is as identical as possible to the correct tag Y.
6. The deep learning-based coal mine gas prediction method of claim 5, wherein the step of deep neural network model DNN training comprises:
establishing an automatic encoder model of a Z layer, defining an activation function of an encoding network as f, outputting the characteristic to be extracted, and performing the encoding process of an automatic encoder of the first layer as follows:
Figure FDA0002275628600000021
Figure FDA0002275628600000022
Figure FDA0002275628600000023
wherein, X0As the input characteristic value, xiThe function f is an activation function for the ith input characteristic attribute; w1As a weight matrix between the input layer and the hidden layer, b1Is an offset;
h is to be1As input to the second hidden layer and train W2And b2The same method is carried out until the weight W of the z-th layer is trainedzAnd bz；
Outputting H of the encoded network through the decoding networkzY is retrieved by the decoding network so that the decoded output Y is as equal as possible to that of the correct tag Y.
7. The deep learning-based coal mine gas prediction method of claim 6, wherein the decoding network utilizes a supervised back propagation algorithm to fine tune network parameters throughout the network. The fine tuning process is done by minimizing the reconstruction error E (θ), and the process of parameter update is as follows:
Figure FDA0002275628600000024
Figure FDA0002275628600000025
wherein M is the number of training set samples, y is the output value of the network, TkIs an offset vector, JθThen the cost function is expressed, theta is the singular stack of the layers of the network, theta is { W, b }, and theta is { theta }1,θ2,…,θzThe backpropagation algorithm updates the stacked theta, α is the learning rate.
CN201911121709.6A 2019-11-15 2019-11-15 Coal mine gas prediction method based on deep learning Pending CN111126658A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201911121709.6A CN111126658A (en) 2019-11-15 2019-11-15 Coal mine gas prediction method based on deep learning

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201911121709.6A CN111126658A (en) 2019-11-15 2019-11-15 Coal mine gas prediction method based on deep learning

Publications (1)

Publication Number Publication Date
CN111126658A true CN111126658A (en) 2020-05-08

Family

ID=70495978

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201911121709.6A Pending CN111126658A (en) 2019-11-15 2019-11-15 Coal mine gas prediction method based on deep learning

Country Status (1)

Country Link
CN (1) CN111126658A (en)

Cited By (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111860645A (en) * 2020-07-20 2020-10-30 中科三清科技有限公司 Method and device for repairing default value in volatile organic compound observation data
CN111860940A (en) * 2020-06-02 2020-10-30 山西潞安环保能源开发股份有限公司 Coal mine distribution network power failure accident prediction method based on logistic regression algorithm
CN112183901A (en) * 2020-11-06 2021-01-05 贵州工程应用技术学院 Coal and gas outburst strength prediction method based on deep learning
CN112712192A (en) * 2020-11-24 2021-04-27 江苏中矿安华科技发展有限公司 Coal mine gas concentration prediction method combining integrated learning and weighted extreme learning machine
CN113485498A (en) * 2021-07-19 2021-10-08 北京工业大学 Indoor environment comfort level adjusting method and system based on deep learning
CN113689032A (en) * 2021-08-09 2021-11-23 陕煤集团神木张家峁矿业有限公司 Multi-sensor fusion gas concentration multi-step prediction method based on deep learning
CN113743486A (en) * 2021-08-23 2021-12-03 北京科技大学 Method for predicting tunneling head coal and gas outburst danger by applying gas concentration after blasting
CN115081156A (en) * 2022-07-21 2022-09-20 太原理工大学 Self-perception, self-decision and self-execution intelligent ventilation control platform and control method for mine
CN116738226A (en) * 2023-05-26 2023-09-12 北京龙软科技股份有限公司 Gas emission quantity prediction method based on self-interpretable attention network
CN116975775A (en) * 2023-06-29 2023-10-31 中信重工开诚智能装备有限公司 Deep learning method for coal mine gas outburst prediction based on multi-mode data fusion

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104156422A (en) * 2014-08-06 2014-11-19 辽宁工程技术大学 Gas concentration real-time prediction method based on dynamic neural network
CN107122861A (en) * 2017-04-28 2017-09-01 辽宁工程技术大学 A kind of Forecast of Gas Emission method based on PCA PSO ELM
CN110135564A (en) * 2019-05-15 2019-08-16 深圳朴生智能科技有限公司 Coal mine gas sensing data method for detecting abnormality based on edge calculations

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104156422A (en) * 2014-08-06 2014-11-19 辽宁工程技术大学 Gas concentration real-time prediction method based on dynamic neural network
CN107122861A (en) * 2017-04-28 2017-09-01 辽宁工程技术大学 A kind of Forecast of Gas Emission method based on PCA PSO ELM
CN110135564A (en) * 2019-05-15 2019-08-16 深圳朴生智能科技有限公司 Coal mine gas sensing data method for detecting abnormality based on edge calculations

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
李兵等: "基于PCA-ELM的矿井瓦斯涌出量预测", 《世界科技研究与发展》 *
米翔繁: "基于PCA-BP神经网络的回采工作面瓦斯涌出量预测研究", 《低碳世界》 *
邓俊锋: "基于自动编码器组合的深度学习优化方法", 《计算机应用》 *

Cited By (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111860940A (en) * 2020-06-02 2020-10-30 山西潞安环保能源开发股份有限公司 Coal mine distribution network power failure accident prediction method based on logistic regression algorithm
CN111860940B (en) * 2020-06-02 2024-03-19 湖北凌云电力建设有限公司 Coal mine power distribution network power failure accident prediction method based on logistic regression algorithm
CN111860645A (en) * 2020-07-20 2020-10-30 中科三清科技有限公司 Method and device for repairing default value in volatile organic compound observation data
CN112183901A (en) * 2020-11-06 2021-01-05 贵州工程应用技术学院 Coal and gas outburst strength prediction method based on deep learning
CN112712192A (en) * 2020-11-24 2021-04-27 江苏中矿安华科技发展有限公司 Coal mine gas concentration prediction method combining integrated learning and weighted extreme learning machine
CN112712192B (en) * 2020-11-24 2021-09-03 江苏中矿安华科技发展有限公司 Coal mine gas concentration prediction method combining integrated learning and weighted extreme learning machine
CN113485498A (en) * 2021-07-19 2021-10-08 北京工业大学 Indoor environment comfort level adjusting method and system based on deep learning
CN113689032A (en) * 2021-08-09 2021-11-23 陕煤集团神木张家峁矿业有限公司 Multi-sensor fusion gas concentration multi-step prediction method based on deep learning
CN113743486B (en) * 2021-08-23 2023-09-29 北京科技大学 Method for predicting heading coal and gas outburst risk by using post-blasting gas concentration
CN113743486A (en) * 2021-08-23 2021-12-03 北京科技大学 Method for predicting tunneling head coal and gas outburst danger by applying gas concentration after blasting
CN115081156A (en) * 2022-07-21 2022-09-20 太原理工大学 Self-perception, self-decision and self-execution intelligent ventilation control platform and control method for mine
CN115081156B (en) * 2022-07-21 2022-11-25 太原理工大学 Self-perception, self-decision and self-execution intelligent ventilation control platform and control method for mine
CN116738226A (en) * 2023-05-26 2023-09-12 北京龙软科技股份有限公司 Gas emission quantity prediction method based on self-interpretable attention network
CN116738226B (en) * 2023-05-26 2024-03-12 北京龙软科技股份有限公司 Gas emission quantity prediction method based on self-interpretable attention network
CN116975775A (en) * 2023-06-29 2023-10-31 中信重工开诚智能装备有限公司 Deep learning method for coal mine gas outburst prediction based on multi-mode data fusion
CN116975775B (en) * 2023-06-29 2024-01-30 中信重工开诚智能装备有限公司 Deep learning method for coal mine gas outburst prediction based on multi-mode data fusion

Similar Documents

Publication Publication Date Title
CN111126658A (en) Coal mine gas prediction method based on deep learning
CN109492822B (en) Air pollutant concentration time-space domain correlation prediction method
CN104657744B (en) A kind of multi-categorizer training method and sorting technique based on non-determined Active Learning
Ren et al. TBSM: A traffic burst-sensitive model for short-term prediction under special events
CN112101480A (en) Multivariate clustering and fused time sequence combined prediction method
CN110570133A (en) building engineering dynamic cost analysis and control system based on big data
CN112329990A (en) User power load prediction method based on LSTM-BP neural network
CN111008726B (en) Class picture conversion method in power load prediction
CN115050177A (en) Expressway congestion prediction method based on space-time multi-head attention mechanism
CN112434848A (en) Nonlinear weighted combination wind power prediction method based on deep belief network
CN110621026A (en) Base station flow multi-time prediction method
CN111950195B (en) Project progress prediction method based on portrait system and depth regression model
CN114066073A (en) Power grid load prediction method
CN113627070A (en) Short-term photovoltaic power prediction method
CN113051811A (en) Multi-mode short-term traffic jam prediction method based on GRU network
CN116187835A (en) Data-driven-based method and system for estimating theoretical line loss interval of transformer area
CN113449919B (en) Power consumption prediction method and system based on feature and trend perception
CN112927507B (en) Traffic flow prediction method based on LSTM-Attention
CN109594967A (en) A kind of lock of tool drilling detection method for early warning based on well logging big data
Wei et al. Processing method of missing data in dam safety monitoring
CN113033898A (en) Electrical load prediction method and system based on K-means clustering and BI-LSTM neural network
CN117131449A (en) Data management-oriented anomaly identification method and system with propagation learning capability
CN117170221A (en) Artificial intelligence control system for sewage treatment
CN107590747A (en) Power grid asset turnover rate computational methods based on the analysis of comprehensive energy big data
CN109635008B (en) Equipment fault detection method based on machine learning

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: 20200508

RJ01 Rejection of invention patent application after publication