CN114000907A - Mine ventilation equipment intelligent regulation and control system based on digital twin technology - Google Patents
Mine ventilation equipment intelligent regulation and control system based on digital twin technology Download PDFInfo
- Publication number
- CN114000907A CN114000907A CN202111509398.8A CN202111509398A CN114000907A CN 114000907 A CN114000907 A CN 114000907A CN 202111509398 A CN202111509398 A CN 202111509398A CN 114000907 A CN114000907 A CN 114000907A
- Authority
- CN
- China
- Prior art keywords
- ventilation
- value
- entity
- data
- control system
- 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
- 238000009423 ventilation Methods 0.000 title claims abstract description 77
- 238000005516 engineering process Methods 0.000 title claims description 8
- 238000000034 method Methods 0.000 claims abstract description 28
- 238000009412 basement excavation Methods 0.000 claims abstract description 13
- 230000003993 interaction Effects 0.000 claims abstract description 5
- 238000012545 processing Methods 0.000 claims abstract description 4
- 239000002245 particle Substances 0.000 claims description 48
- 238000013528 artificial neural network Methods 0.000 claims description 34
- 230000006870 function Effects 0.000 claims description 29
- 239000007789 gas Substances 0.000 claims description 26
- 230000002457 bidirectional effect Effects 0.000 claims description 15
- 230000008569 process Effects 0.000 claims description 15
- 238000012549 training Methods 0.000 claims description 15
- 239000000428 dust Substances 0.000 claims description 13
- 239000003245 coal Substances 0.000 claims description 11
- 238000005065 mining Methods 0.000 claims description 10
- 230000004913 activation Effects 0.000 claims description 7
- 230000003044 adaptive effect Effects 0.000 claims description 6
- 238000007781 pre-processing Methods 0.000 claims description 5
- 238000012360 testing method Methods 0.000 claims description 5
- 230000001105 regulatory effect Effects 0.000 claims description 4
- 238000004088 simulation Methods 0.000 claims description 4
- 230000009191 jumping Effects 0.000 claims description 3
- 230000008447 perception Effects 0.000 claims description 3
- 230000000630 rising effect Effects 0.000 claims description 3
- UGFAIRIUMAVXCW-UHFFFAOYSA-N Carbon monoxide Chemical compound [O+]#[C-] UGFAIRIUMAVXCW-UHFFFAOYSA-N 0.000 claims description 2
- 229910002091 carbon monoxide Inorganic materials 0.000 claims description 2
- 230000001276 controlling effect Effects 0.000 claims description 2
- 238000012795 verification Methods 0.000 claims description 2
- 238000004891 communication Methods 0.000 claims 1
- 238000007418 data mining Methods 0.000 claims 1
- 230000000007 visual effect Effects 0.000 claims 1
- 210000004027 cell Anatomy 0.000 description 16
- 230000008859 change Effects 0.000 description 4
- 238000013507 mapping Methods 0.000 description 4
- 210000002569 neuron Anatomy 0.000 description 4
- 230000002441 reversible effect Effects 0.000 description 4
- 238000003062 neural network model Methods 0.000 description 3
- RAHZWNYVWXNFOC-UHFFFAOYSA-N Sulphur dioxide Chemical compound O=S=O RAHZWNYVWXNFOC-UHFFFAOYSA-N 0.000 description 2
- 230000001133 acceleration Effects 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 2
- 230000001413 cellular effect Effects 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000004927 fusion Effects 0.000 description 2
- 230000007787 long-term memory Effects 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 230000000306 recurrent effect Effects 0.000 description 2
- 230000000717 retained effect Effects 0.000 description 2
- 238000013058 risk prediction model Methods 0.000 description 2
- 230000006403 short-term memory Effects 0.000 description 2
- 238000012546 transfer Methods 0.000 description 2
- 230000005641 tunneling Effects 0.000 description 2
- MGWGWNFMUOTEHG-UHFFFAOYSA-N 4-(3,5-dimethylphenyl)-1,3-thiazol-2-amine Chemical compound CC1=CC(C)=CC(C=2N=C(N)SC=2)=C1 MGWGWNFMUOTEHG-UHFFFAOYSA-N 0.000 description 1
- RWSOTUBLDIXVET-UHFFFAOYSA-N Dihydrogen sulfide Chemical compound S RWSOTUBLDIXVET-UHFFFAOYSA-N 0.000 description 1
- 206010063385 Intellectualisation Diseases 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 description 1
- 230000004888 barrier function Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000007664 blowing Methods 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000004880 explosion Methods 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 229910000037 hydrogen sulfide Inorganic materials 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- JCXJVPUVTGWSNB-UHFFFAOYSA-N nitrogen dioxide Inorganic materials O=[N]=O JCXJVPUVTGWSNB-UHFFFAOYSA-N 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 229910052760 oxygen Inorganic materials 0.000 description 1
- 239000001301 oxygen Substances 0.000 description 1
- 231100000614 poison Toxicity 0.000 description 1
- 230000007096 poisonous effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 239000011378 shotcrete Substances 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Images
Classifications
-
- E—FIXED CONSTRUCTIONS
- E21—EARTH DRILLING; MINING
- E21F—SAFETY DEVICES, TRANSPORT, FILLING-UP, RESCUE, VENTILATION, OR DRAINING IN OR OF MINES OR TUNNELS
- E21F1/00—Ventilation of mines or tunnels; Distribution of ventilating currents
- E21F1/006—Ventilation at the working face of galleries or tunnels
-
- E—FIXED CONSTRUCTIONS
- E21—EARTH DRILLING; MINING
- E21F—SAFETY DEVICES, TRANSPORT, FILLING-UP, RESCUE, VENTILATION, OR DRAINING IN OR OF MINES OR TUNNELS
- E21F17/00—Methods or devices for use in mines or tunnels, not covered elsewhere
- E21F17/18—Special adaptations of signalling or alarm devices
Abstract
The invention provides an intelligent regulation and control system and method of mine ventilation equipment based on digital twins, wherein the system comprises a physical entity of a digital twins system, a virtual entity, twins data and an intelligent regulation and control system of the ventilation equipment; the physical entity of the digital twin system comprises a mine fully-mechanized excavation scene, which refers to an excavation working face, physical facilities of a ventilation system and various sensors; the digital twin system virtual entity refers to a virtual modeling environment for a physical entity, truly maps data information of the physical entity and provides a virtual debugging environment for the ventilation intelligent regulation and control system; the intelligent regulation and control system runs on the virtual entity and acts on the physical entity, and is a key module for realizing real-time interaction between the virtual entity and the physical entity; the intelligent regulation and control system of the ventilation system is responsible for processing twin data, and a decision scheme is generated through data information to provide intelligent decision information for ventilation system equipment.
Description
Technical Field
The invention relates to the technical field of mine ventilation equipment regulation and control, in particular to an intelligent regulation and control system of mine ventilation equipment based on a digital twin technology.
Technical Field
Coal mine intellectualization is a core technical support for high-quality development of the coal industry, coal occupies an important position in the energy structure of China, and safe and efficient mining is a always pursued target in the coal field. The mine ventilation system can supply enough fresh air underground to meet the requirement of personnel on oxygen; poisonous and harmful gas and dust in the well are diluted, and the safe production is ensured; the underground climate is adjusted, and a good working environment is created. In the coal mining process, a large amount of harmful factors such as gas, dust and the like can be accumulated on the fully-mechanized excavating surface, and the life health condition of mining personnel is seriously influenced. How to efficiently and conveniently reduce the content of harmful gas in a roadway and control the content of the harmful gas in a safety range is a problem which needs to be solved in the coal mining process.
At present, a fixed ventilator is used for blowing and exhausting air to dilute gas and dust in a roadway in a mine, the original ventilation total amount control mode is low in ventilation efficiency, the working state of an air outlet of ventilation equipment cannot be dynamically adjusted in real time along with the operation environment of the roadway, when the concentration of harmful gas in the mine is abnormal, the ventilation system cannot be well sensed and adjusted, and the problems that the ventilation equipment is not adjusted in place and is not timely and the like can cause huge personnel and property loss at a critical moment. Therefore, the research on the intelligent regulation and control method of the ventilation system under the mine has important practical significance for avoiding property loss and casualties under the mine and fundamentally solving the problem of coal mine safety.
Disclosure of Invention
In order to improve the technical problem of ventilation, the invention aims to provide an intelligent regulation and control system of mine ventilation equipment based on a digital twin technology, so that the optimized cooperative work of a ventilation system in a coal mine is realized, and an effective solution is provided for the optimization of a mine ventilation environment.
In order to achieve the purpose, the invention adopts the technical scheme that:
the utility model provides a mine ventilation equipment intelligent regulation and control system based on digital twin technique, includes ventilation equipment intelligent regulation and control system physical entity, ventilation equipment intelligent regulation and control system virtual entity, ventilation equipment intelligent decision-making system and digital twin data, wherein:
the physical entity of the intelligent ventilation equipment regulation and control system refers to an actual coal mine fully-mechanized excavation scene, and particularly relates to mining mechanical equipment, ventilation equipment, intelligent regulation and control equipment and various air quality detection sensors in the fully-mechanized excavation process.
The intelligent ventilation regulating and controlling equipment can regulate and control the state parameters such as the position parameter, the air outlet size, the angle deflection, the air outlet time length and the like of the ventilator around the tunneling working face through a programmable controller together with various sensors.
The various sensors in the physical scene include wind speed sensors, gas concentration sensors, dust concentration sensors, distance measuring sensors, CO sensors, temperature sensors, and the like. The data acquisition device sends the data that all kinds of sensors gathered to the information collection point, and the information collection point unifies data fusion arrangement to information such as sensor data and equipment data, realizes the real-time collection and the perception of each position temperature, wind speed, dust concentration, gas concentration, CO concentration condition in tunnelling working face and the tunnelling working channel.
The virtual entity of the intelligent regulation and control system of the ventilation equipment refers to the real mapping of a physical entity in a computer, the virtual entity accurately maps all elements of the physical entity in real time, can provide a virtual modeling environment for a digital twin system, provides a running environment for data flow, provides an intelligent decision running platform based on deep learning for the digital twin system, the virtual entity and the physical entity are connected in real time through various sensor data in a model mode, the virtual entity and the physical entity acquire real-time state data of the physical entity based on an open platform unified framework protocol of a programmable controller, and a mapping relation is established with simulation data formed by the physical entity by calling a corresponding model in the data driving virtual entity. Meanwhile, the control data information intelligently regulated and controlled by the ventilation equipment is transmitted by adopting an open platform unified frame protocol, the position parameters, the air outlet size, the angle deflection, the air outlet time length and the like of the ventilator are controlled in real time, and the real-time sensing and interaction effect between the physical entity and the virtual entity can be realized by continuously optimizing the acquisition process in an iterative manner.
The ventilation intelligent decision system is responsible for processing twin data, generating decision information by fusing the data, providing an intelligent regulation and control scheme for a ventilation equipment regulation and control device, and obtaining iterative optimization of regulation and control parameters of the ventilation equipment;
the twin data refers to the real-time state information of the tunneling working face, the underground facilities, such as azimuth temperature, wind speed, dust concentration, gas concentration, CO concentration and the like acquired by the sensor in real time and the optimized regulation and control data obtained by prediction and regulation in the intelligent regulation and control system of the ventilation equipment.
The prediction function of the intelligent ventilation equipment regulation and control system is to adopt a bidirectional LSTM neural network model to carry out mining analysis on digital twin data flowing in the wind direction, extract potential features of the data, predict the change trend of the concentrations of gas, dust and CO of a fully-mechanized excavation face and a mining channel in a mine at a future moment, if the rising trend of the concentrations of the gas, the dust and the CO is predicted in the future time, the intelligent ventilation equipment regulation and control system starts an early warning function to send out early warning information to a terminal panel, simultaneously adopts a PSO-BP neural network to generate an intelligent ventilation equipment regulation and control scheme, analyzes, adjusts and optimizes parameter information of various ventilation equipment such as an air duct, an exhaust fan and the like, carries out simulation and verification on the generated intelligent ventilation equipment regulation and control scheme of the PSO-BP neural network on the digital twin system, and carries out real-time adjustment on the ventilation equipment after the feasibility is simulated and verified, so as to follow up the change of the operation environment in the mine and ensure the safety of the environment.
2. A harmful gas early warning method of an intelligent regulation and control system of ventilation equipment comprises the following steps:
s1, collecting data of each gas sensor in the mine and parameter information of ventilation equipment to obtain an original data set, and preprocessing the data to form a data set for neural network training;
s2, constructing a bidirectional LSTM neural network fault prediction model, training the bidirectional LSTM neural network by using the data set obtained in the S1, updating model parameters, and outputting the trained neural network;
s3, performing danger prediction on real-time data information in the digital twin system by using the bidirectional LSTM neural network trained in the S2, and when the prediction result is that danger will occur, early warning on an interactive page by the digital twin system;
and step S4, when the early warning is sent out in the step S3, the digital twin system calls a PSO-BP neural network module to generate an emergency scheme, the PSO-BP neural network takes the twin data of the current sensor and the ventilation equipment as an input layer, and the neural network outputs various regulation and control parameters of the ventilation equipment as equipment regulation and control indexes.
In step S1, the method for preprocessing the original data set includes dividing the training set and the test set, normalizing the data, and setting an alarm tag for the data.
The data set is divided by dividing 90% of the data set into training set and the remaining 10% of the data set into testing set.
The data normalization method comprises the following steps:
in the formula, XnormalFor normalized standard data, XiIs the original data, XminIs the minimum value, X, in the data setmaxIs the maximum value in the data set.
The method for setting the labels for the data is that according to the past alarm records, the multi-level alarm labels are added to the historical sensor data and the equipment parameter data at the same time point, and the alarm labels with different levels are divided according to the severity of the alarm condition.
The method for constructing the bidirectional LSTM neural network fault prediction model in the step S2 comprises the following steps:
the LSTM recurrent neural network learns the dependency relationship in a long time window through a door structure, short-term memory is realized through an activation function in the network, and updating of the weight is used for long-term memory. Wherein C ist-1Cell status of the previous step, ht-1Output value, x, for previous hidden statetIs the input to the current LSTM unit. CtFor renewed cell state, htIs the output value of the current hidden state. The core of the LSTM unit is a cell state, the cell state transmits related information along a time sequence, and the updating of the state is determined by a forgetting gate, an input gate and an output gate.
Firstly, constructing a unidirectional LSTM unit, wherein:
forget door ftThe decision on how much information should be discarded or retained can be expressed as:
ft＝σ(Wf·[ht-1,xt]+bf)
in the formula: f. oftPassing from previous hidden state information and current input information to sigmoid function, WfIs a weight term; bfIs the bias term.
Input door itIt is determined how to update the cell state, i.e. to selectively record new information into the cell state. Can be expressed as:
it＝σ(Wi·[ht-1,xt]+bi)
in the formula: h ist-1And xtTransferring the cell state to a function and simultaneously transmitting the cell state to a tanh activation function to obtain a cell state candidate value
Both determine a novel cellular state Ct。
Output gate otDetermining the output of the current hidden state, similar to the input gate, htFrom otAnd CtThe carried information is determined. Can be expressed as:
ot＝σ(Wo·[ht-1,xt]+bo)
ht＝ot*tanh(Ct)
and then constructing a bidirectional LSTM model, wherein the bidirectional LSTM predicts the output based on the whole time sequence, firstly divides the hidden layer neuron into 2 parts in a positive time direction and a negative time direction, has 2 independent hidden layers, and then feeds forward to the same output layer, and simultaneously comprises the past and future sequence information. The layer 1 LSTM calculates the current time point sequence information, the layer 2 LSTM reads the same sequence in reverse direction, adds the reverse sequence information, and each layer of LSTM has different parameters.
The risk prediction model process is as follows:
1. firstly, preprocessing a data set, and adding danger warning labels to a training set and a test set;
2. the input of the network is the length of the time series and the characteristic number, and the corresponding time window size is set.
3. The time series of each sensor and device parameter is divided according to the time window size, and the normalized data in a two-dimensional format is used directly as the model input.
4. And adjusting the weight parameters of the neural network according to the average loss function of each batch, and minimizing the loss function value.
5. And setting the output of the full connection layer as the dimension, the activation function and the dropout probability of the label.
6. And repeating the steps, and finishing the model training when the loss function value reaches a threshold value or reaches the maximum iteration times.
7. And training the early warning level marked by the sample to be output as a target of the model.
The PSO-BP neural network regulation and control model and the construction method in the step S4 are as follows:
the BP neural network generally adopts a 3-layer structure, and the basic formula structure is shown as the following formula (1):
y＝purelin(Wl2*tansigWl1*xn+θ1)+θ2 (1)
wherein tansig () is the hyperbolic tangent S function between the input layer to the hidden layer; puerlin () is a linear transfer function between the hidden layer to the output layer; y is the output vector of the BP network; x is the number ofnIs an input vector; wl1The connection weight from the input layer to the hidden layer; wl2The connection weight between the hidden layer and the output layer; theta1For input layer to hidden layer connection threshold；θ2Is the connection threshold of the hidden layer to the output layer.
3. A harmful gas early warning method of an intelligent regulation and control system of ventilation equipment comprises the following steps:
the PSO algorithm is to iteratively find the optimal solution among a group of random particles. In each iteration process, the particles are updated by self through two 'extrema' of the optimal solution Pbest searched by the particles and the global extremum Gbest searched by the whole particle group. A particle group consisting of n particles arranged in a D-dimensional search vector, wherein the position of the ith particle is Xi＝(Xi1,Xi2,…,XiD) Flying speed is Vi＝(Vi1,Vi2,…,ViD) I ∈ 1,2, …, n. The optimum position searched is Pi＝(Pi1,Pi2,…,PiD) I ∈ 1,2, …, n. In the global particle swarm, the optimal solution positions recorded by all Pi are Pg＝(Pg1,Pg2,…,PgD). During each iteration, the particle state is updated according to the following formula:
Vij(t+1)＝ω(t)Vij(t)+c1*rand()*(Pij(t)-Xij(t))+c2*rand()*(Pgj(t)-Xij(t))(2)
Xij(t+1)＝Xij+Vij(t+1) (3)
ω(t)＝ωmax-t(ωmax-ωmin)/tmax (4)
wherein j ∈ 1,2, …, d; c. C1And c2Is the learning rate, i.e., acceleration constant; range () is [0,1 ]]A random number function; t is the number of iterations, tmaxIs the maximum number of iterations; omegaminIs the minimum weight, ωmaxIs the maximum weight.
When the PSO algorithm is used for optimizing the BP network, the position vector X of the particle is the weight W and the threshold value theta of the BP network. In the particle search process, the mean square error shown in the following formula (5) is used, that is: adaptation value FitnessiTo a minimum.
Wherein N is the number of training set samples; m is the number of output network neurons;
i is the ideal state value of the jth output network node of the ith sample; wj,iIs the actual state value of the jth output network node for the ith sample.
The method comprises the following steps of constructing a PSO-BP neural network regulation decision model:
1. constructing a PSO-BP network with a 3-layer structure and initializing;
2. initializing PSO including the number of particles, the speed and the position of the particle swarm and the like;
3. calculating the Pbest value of each particle, and selecting the best Pbest value as the Gbest value;
4. and (5) calculating an adaptive value of each particle according to the formula (5), and if the adaptive value is better than the current Pbest value of the particle, updating the position of the particle and the Pbest value. If the Pbest values of all the particles are superior to the current Gbest value, updating the Gbest value;
5. updating the weight value according to a formula (4), and updating the particle position and the particle speed according to a formula (2) and a formula (3);
6. and judging whether the maximum iteration times is reached, if so, stopping iteration to obtain the optimal weight W and the threshold theta of the BP network, and otherwise, jumping to the step 3.
The invention has the beneficial effects that:
according to the invention, by constructing the digital twin body of the underground comprehensive excavation face and the roadway, the digitization of the ventilation system in the roadway is realized, so that the physical entity and the virtual entity of the digital twin system are interacted, and the mining scene is effectively reproduced, monitored in real time and accurately controlled; in addition, a data set is constructed through sensors and equipment parameters, a deep learning model is constructed, and the capabilities of autonomous learning, autonomous prediction and autonomous decision making are realized for a digital twin system by utilizing harmful gas concentration early warning based on a bidirectional LSTM neural network and a ventilation equipment regulation and control decision making technology based on a PSO-BP neural network.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained based on these drawings without inventive labor.
FIG. 1 is a view showing a construction of a ventilation digital twin system according to the present invention
FIG. 2 is a flow chart of intelligent regulation and control of a ventilation system according to the present invention
FIG. 3 is a flow chart of a bidirectional LSTM neural network prediction model in the intelligent control system of the present invention
FIG. 4 is a PSO-BP based neural network device regulation model in the intelligent regulation system of the present invention
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
As shown in fig. 1, fig. 1 is a structural diagram of a digital twin system of a ventilation system in the invention, and the structure of the digital twin system comprises a physical entity, a virtual entity, digital twin data and a ventilation regulation intelligent decision system.
The physical entities refer to actual coal mine fully-mechanized excavation scenes, and particularly refer to mining mechanical equipment, ventilation equipment and various air quality detection sensors under a mine in the fully-mechanized excavation process.
The excavating equipment comprises a heading machine, a conveyor, a fan, a water pump, a bolting machine, a shotcrete machine, a pneumatic pick and the like, the ventilation equipment comprises a wind tunnel, an air bridge, an air window, a wind barrier, an air clamp and other physical facilities for guiding wind flow and separating the wind flow by an explosion door, a wind-break wall, an air door and the like, and the various sensors comprise a wind speed sensor, a distance measuring sensor, a temperature sensor and various concentration sensors of harmful gases such as dust, gas, carbon monoxide, hydrogen sulfide, sulfur dioxide and nitrogen dioxide.
The data acquisition device sends the data that all kinds of sensors gathered to the information collection point, realizes the real-time collection and the perception of each position temperature, wind speed and harmful gas concentration condition in tunnelling working face and the mine working channel.
The virtual entity of the intelligent regulation and control system of the ventilation equipment refers to the real mapping of a physical entity in a computer, the virtual entity accurately maps all elements of the physical entity in real time, can provide a virtual modeling environment for a digital twin system, provides a running environment for data flow, provides an intelligent decision running platform based on deep learning for the digital twin system, the virtual entity and the physical entity are connected in real time through various sensor data in a model mode, the virtual entity and the physical entity acquire real-time state data of the physical entity based on an open platform unified framework protocol of a programmable controller, and a mapping relation is established with simulation data formed by the physical entity by calling a corresponding model in the data driving virtual entity. Meanwhile, the control data information intelligently regulated and controlled by the ventilation equipment is transmitted by adopting an open platform unified frame protocol, the position parameters, the air outlet size, the angle deflection, the air outlet time length and the like of the ventilator are controlled in real time, and the real-time sensing and interaction effect between the physical entity and the virtual entity can be realized by continuously optimizing the acquisition process in an iterative manner.
The ventilation regulation and control intelligent decision system refers to a module related to ventilation equipment and ventilation regulation and control in a digital twin virtual entity, twin data are generated after data information generated by various sensors is subjected to data fusion, a decision scheme is generated through deep neural network processing, and an intelligent regulation and control method is provided for an intelligent regulation and control device of the ventilation equipment.
The twin data refer to real-time state information of the heading face such as azimuth temperature, wind speed, dust concentration, gas concentration and CO concentration acquired by the sensor in real time and optimized regulation data obtained by prediction and regulation in the intelligent regulation and control system of the ventilation equipment.
As shown in fig. 2, fig. 2 is a flow chart of the intelligent regulation and control of the ventilation system in the present invention. The intelligent control system of the ventilation equipment adopts a bidirectional LSTM neural network model to predict the change tendency of the concentration of gas, dust and CO on the fully-mechanized excavation face and the excavation channel in the mine at the future time. If the concentration of harmful gas in the existing fully-mechanized excavation face and the roadway exceeds the standard, an existing emergency treatment standard program is adopted to send out an alarm and take emergency measures; if the concentration of the gas does not exceed the warning standard at present, but the rising trend of the concentrations of gas, dust and CO is predicted in the future time through a neural network model, an intelligent regulation and control system of the ventilation equipment starts a warning function and sends warning information to a terminal panel, meanwhile, a PSO-BP neural network generation equipment intelligent regulation and control scheme is adopted to analyze, adjust and optimize parameter information of various ventilation equipment such as an air duct, an exhaust fan and the like, the intelligent regulation and control scheme generated by the PSO-BP neural network is simulated and verified on a digital twin system, and the ventilation equipment is adjusted in real time after the feasibility is simulated and verified so as to follow the change of the working environment in a mine and ensure the safety of the environment.
FIG. 3 is a flow chart of a bidirectional LSTM neural network prediction model in the intelligent regulation system of the present invention.
The risk prediction model process is as follows:
1. firstly, preprocessing a data set, and adding danger warning labels to a training set and a test set;
2. the input of the network is the length of the time series and the characteristic number, and the corresponding time window size is set.
3. The time series of each sensor and device parameter is divided according to the time window size, and the normalized data in a two-dimensional format is used directly as the model input.
4. And adjusting the weight parameters of the neural network according to the average loss function of each batch, and minimizing the loss function value.
5. And setting the output of the full connection layer as the dimension, the activation function and the dropout probability of the label.
6. And repeating the steps, and finishing the model training when the loss function value reaches a threshold value or reaches the maximum iteration times.
7. And training the early warning level marked by the sample to be output as a target of the model.
The LSTM recurrent neural network learns the dependency relationship in a long time window through a door structure, short-term memory is realized through an activation function in the network, and updating of the weight is used for long-term memory. Wherein C ist-1Cell status of the previous step, ht-1Output value, x, for previous hidden statetIs the input to the current LSTM unit. CtFor renewed cell state, htIs the output value of the current hidden state. The core of the LSTM unit is a cell state, the cell state transmits related information along a time sequence, and the updating of the state is determined by a forgetting gate, an input gate and an output gate.
Firstly, constructing a unidirectional LSTM unit, wherein:
forget door ftThe decision on how much information should be discarded or retained can be expressed as:
ft＝σ(Wf[ht-1,xt]+bf)
in the formula: f. oftPassing from previous hidden state information and current input information to sigmoid function, WfIs a weight term; bfIs the bias term.
Input door itIt is determined how to update the cell state, i.e. to selectively record new information into the cell state. Can be expressed as:
it＝σ(Wi·[ht-1,xt]+bi)
in the formula: h ist-1And xtTransferring the cell state to a function and simultaneously transmitting the cell state to a tanh activation function to obtain a cell state candidate value
Both determine a novel cellular state Ct。
Output gate otDetermining the output of the current hidden state, similar to the input gate, htFrom otAnd CtThe carried information is determined. Can be expressed as:
ot＝σ(Wo·[ht-1,xt]+bo)
ht＝ot*tanh(Ct)
and then constructing a bidirectional LSTM model, wherein the bidirectional LSTM predicts the output based on the whole time sequence, firstly divides the hidden layer neuron into 2 parts in a positive time direction and a negative time direction, has 2 independent hidden layers, and then feeds forward to the same output layer, and simultaneously comprises the past and future sequence information. The layer 1 LSTM calculates the current time point sequence information, the layer 2 LSTM reads the same sequence in reverse direction, adds the reverse sequence information, and each layer of LSTM has different parameters.
FIG. 4 is a PSO-BP based neural network device regulation model in the intelligent regulation system of the present invention
The BP neural network generally adopts a 3-layer structure, and the basic formula structure is shown as the following formula (1):
y＝purelin(Wl2*tansigWl1*xn+θ1)+θ2 (1)
wherein tansig () is the hyperbolic tangent S function between the input layer to the hidden layer; puerlin () is a linear transfer function between the hidden layer to the output layer; y is the output vector of the BP network; x is the number ofnIs an input vector; wl1The connection weight from the input layer to the hidden layer; wl2The connection weight between the hidden layer and the output layer; theta1A connection threshold value from the input layer to the hidden layer; theta2Is the connection threshold of the hidden layer to the output layer.
The PSO algorithm is to iteratively find the optimal solution among a group of random particles. In each iteration process, the particles are updated by self through two 'extrema' of the optimal solution Pbest searched by the particles and the global extremum Gbest searched by the whole particle group. A particle group consisting of n particles arranged in a D-dimensional search vector, wherein the position of the ith particle is Xi＝(Xi1,Xi2,…,XiD) Flying speed is Vi＝(Vi1,Vi2,…,ViD) I ∈ 1,2, …, n. The optimum position searched is Pi＝(Pi1,Pi2,…,PiD) I ∈ 1,2, …, n. In the global particle swarm, the optimal solution positions recorded by all Pi are Pg＝(Pg1,Pg2,…,PgD). During each iteration, the particle state is updated according to the following formula:
Vij(t+1)＝ω(t)Vij(t)+c1*rand()*(Pij(t)-Xij(t))+c2*rand()*(Pgj(t)-Xij(t))(2)
Xij(t+1)＝Xij+Vij(t+1) (3)
ω(t)＝ωmax-t(ωmax-ωmin)/tmax (4)
wherein j ∈ 1,2, …, d; c. C1And c2Is the learning rate, i.e., acceleration constant; range () is [0,1 ]]A random number function; t is the number of iterations, tmaxIs the maximum number of iterations; omegaminIs the minimum weight, ωmaxIs the maximum weight.
When the PSO algorithm is used for optimizing the BP network, the position vector X of the particle is the weight W and the threshold value theta of the BP network. In the particle search process, the mean square error shown in the following formula (5) is used, that is: adaptation value FitnessiTo a minimum.
Wherein N is the number of training set samples; m is the number of output network neurons;
i is the ideal state value of the jth output network node of the ith sample; wj,iIs the actual state value of the jth output network node for the ith sample.
The method comprises the following steps of constructing a PSO-BP neural network regulation decision model:
1. constructing a PSO-BP network with a 3-layer structure and initializing;
2. initializing PSO including the number of particles, the speed and the position of the particle swarm and the like;
3. calculating the Pbest value of each particle, and selecting the best Pbest value as the Gbest value;
4. and (5) calculating an adaptive value of each particle according to the formula (5), and if the adaptive value is better than the current Pbest value of the particle, updating the position of the particle and the Pbest value. If the Pbest values of all the particles are superior to the current Gbest value, updating the Gbest value;
5. updating the weight value according to a formula (4), and updating the particle position and the particle speed according to a formula (2) and a formula (3);
6. and judging whether the maximum iteration times is reached, if so, stopping iteration to obtain the optimal weight W and the threshold theta of the BP network, and otherwise, jumping to the step 3.
Claims (3)
1. The utility model provides a mine ventilation intelligent regulation and control system based on digital twin technique which characterized in that, ventilation system full key element physical entity, ventilation system digital twin virtual entity, ventilation intelligent regulation and control system, wherein:
the physical entity refers to a physical scene formed by all elements related to an underground ventilation system in a coal mine excavation process, and comprises related physical elements including physical comprehensive excavation equipment, various ventilation equipment and sensors in a mining roadway.
The digital twin virtual entity refers to a visual environment established by modeling of information such as data, parameters and the like generated by the ventilation system full-element physical entity; the virtual debugging environment is used for providing environment simulation, virtual debugging and decision verification for the digital twin system, and the virtual entity is used for providing a virtual environment for autonomous learning, autonomous prediction and autonomous decision for the digital twin system; the virtual entity establishes real-time perception and real-time synchronization with the physical entity through a sensing communication technology, and controls the physical entity through an intelligent regulation and control system, so that real-time interaction and real-time control are realized.
The intelligent regulation and control system runs on the virtual entity and acts on the physical entity, and is a key module for realizing real-time interaction between the virtual entity and the physical entity; the intelligent regulation and control system of the ventilation system is responsible for processing twin data, and a decision scheme is generated through data information to provide intelligent decision information for ventilation system equipment.
2. The intelligent mine ventilation regulating and controlling system based on the digital twin technology as claimed in claim 1, wherein a bidirectional LSTM deep neural network is adopted in the intelligent decision making system to perform data mining on equipment parameters and sensor parameters in a mine roadway, data characteristics are learned so as to predict the concentration of harmful gases such as gas, dust and carbon monoxide at a future moment, and if the concentration of the harmful gases at the future moment has a rising trend, the intelligent decision making system can send out an early warning signal and further process the early warning signal.
The bidirectional LSTM deep neural network risk early warning model process comprises the following steps:
step1, firstly, preprocessing a data set, and adding danger warning labels to a training set and a testing set;
step2, the input of the network is the time sequence length and the feature number, and the corresponding time window size is set.
Step3 time series of each sensor and device parameter are divided according to time window size, normalized data in two-dimensional format is used directly as model input.
Step4, according to the average loss function of each batch, the weight parameter of the neural network is adjusted, and the loss function value is minimized.
Step5, set the output of the fully-connected layer as the dimensions of the label, the activation function and the dropout probability.
And Step6, repeating the steps, and finishing the model training when the loss function value reaches a threshold value or reaches the maximum iteration number.
Step7 training the early warning level of the sample label as the target output of the model.
3. The intelligent regulation and control system for mine ventilation based on digital twin technology as claimed in claim 1, wherein a PSO-BP neural network is used to generate an intelligent regulation and control scheme for ventilation equipment in a mine, and the main contents of the intelligent regulation and control scheme comprise:
the method comprises the following steps of constructing a PSO-BP neural network regulation decision model:
step1, constructing a PSO-BP network with a 3-layer structure and initializing;
step2, initializing PSO including the particle number, the speed and the position of the particle group and the like;
step3, calculating the Pbest value of each particle, and selecting the best Pbest value as the Gbest value;
and Step4, calculating the adaptive value of each particle, and if the adaptive value is better than the current Pbest value of the particle, updating the position of the particle and the Pbest value. If the Pbest values of all the particles are superior to the current Gbest value, updating the Gbest value;
step5, updating the weight value, and updating the position and the speed of the particles according to a formula;
and Step6, judging whether the maximum iteration times is reached, if so, stopping iteration to obtain the optimal weight W and the threshold value theta of the BP network, otherwise, jumping to the Step 3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111509398.8A CN114000907A (en) | 2021-12-10 | 2021-12-10 | Mine ventilation equipment intelligent regulation and control system based on digital twin technology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111509398.8A CN114000907A (en) | 2021-12-10 | 2021-12-10 | Mine ventilation equipment intelligent regulation and control system based on digital twin technology |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114000907A true CN114000907A (en) | 2022-02-01 |
Family
ID=79931858
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111509398.8A Pending CN114000907A (en) | 2021-12-10 | 2021-12-10 | Mine ventilation equipment intelligent regulation and control system based on digital twin technology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114000907A (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114484822A (en) * | 2022-02-10 | 2022-05-13 | 中国海洋大学 | Ocean platform ventilation system control method based on temperature and hydrogen sulfide concentration control |
| CN114738031A (en) * | 2022-04-15 | 2022-07-12 | 北京科技大学 | Natural and mechanical ventilation coupled construction tunnel digital twin intelligent ventilation method |
| CN115027641A (en) * | 2022-06-10 | 2022-09-09 | 江南造船(集团)有限责任公司 | Ship ventilation management method and system based on virtual-real interaction |
| CN115184563A (en) * | 2022-09-08 | 2022-10-14 | 北京中环高科环境治理有限公司 | Chemical workshop field data acquisition method based on digital twinning |
| CN115270532A (en) * | 2022-09-29 | 2022-11-01 | 山东新矿信息技术有限公司 | Data processing method and system based on mine intellectualization |
| CN115358494A (en) * | 2022-10-20 | 2022-11-18 | 中电建铁路建设投资集团有限公司 | Danger early warning method for subway shield underpass construction |
| CN115587542A (en) * | 2022-11-02 | 2023-01-10 | 四川奥凸水处理系统工程有限公司 | Underground water inversion simulation method, system, equipment and medium based on reinforcement learning |
| CN115788591A (en) * | 2023-02-08 | 2023-03-14 | 山东大学 | Monitoring and intelligent regulation and control system for dust and harmful gas in deep-buried long tunnel |
| CN116090076A (en) * | 2023-03-07 | 2023-05-09 | 四川省公路规划勘察设计研究院有限公司 | Gabion abutment building system under complex environment and rapid building method thereof |
| RU2798530C1 (en) * | 2022-05-04 | 2023-06-23 | федеральное государственное автономное образовательное учреждение высшего образования "Пермский национальный исследовательский политехнический университет" | System for controlling processes in an underground mining enterprise based on demand for electricity |
| CN117411383A (en) * | 2023-12-15 | 2024-01-16 | 华夏天信智能物联股份有限公司 | Integrated mining variable-frequency speed regulation equipment control method and device and electronic equipment |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2004169341A (en) * | 2002-11-19 | 2004-06-17 | Hitachi Ltd | Method and apparatus for controlling tunnel ventilation |
| CN103716324A (en) * | 2013-12-31 | 2014-04-09 | 重庆邮电大学 | Virtual mine risk-taking behavior implementation system and method based on multiple agents |
| CN106777528A (en) * | 2016-11-25 | 2017-05-31 | 山东蓝光软件有限公司 | The holographic forecast method of mine air-required volume |
| WO2017176944A1 (en) * | 2016-04-05 | 2017-10-12 | Fractal Industries, Inc. | System for fully integrated capture, and analysis of business information resulting in predictive decision making and simulation |
| CN108133289A (en) * | 2017-12-21 | 2018-06-08 | 中国铁建电气化局集团有限公司 | Tunnel ventilation control method and system based on environmental forecasting |
| CN110633540A (en) * | 2019-09-24 | 2019-12-31 | 青岛理工大学 | Metal mine ventilation three-dimensional visual simulation aided decision control system and method |
| US20200348038A1 (en) * | 2019-07-12 | 2020-11-05 | Johnson Controls Technology Company | Hvac system design and operational tool for building infection control |
| CN112684694A (en) * | 2020-11-15 | 2021-04-20 | 杭州哲达科技股份有限公司 | Blast furnace blast and TRT real-time monitoring and simulation control system based on digital twin body |
| CN112761700A (en) * | 2021-01-14 | 2021-05-07 | 成都茹附敬科技有限公司 | Training method of neural network for wind gear control of underground wind adjusting device |
| CN113356916A (en) * | 2021-07-08 | 2021-09-07 | 长安大学 | Mine air flow regulation and control virtual system based on digital twin technology and intelligent regulation and control method |
| CN113530589A (en) * | 2021-07-29 | 2021-10-22 | 西安重装韩城煤矿机械有限公司 | Intelligent local ventilation system and method for supplying air to coal mine driving face according to needs |
| CN116576527A (en) * | 2023-05-19 | 2023-08-11 | 国核电力规划设计研究院有限公司 | Ventilation method, ventilation device, electronic equipment and storage medium |
-
2021
- 2021-12-10 CN CN202111509398.8A patent/CN114000907A/en active Pending
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2004169341A (en) * | 2002-11-19 | 2004-06-17 | Hitachi Ltd | Method and apparatus for controlling tunnel ventilation |
| CN103716324A (en) * | 2013-12-31 | 2014-04-09 | 重庆邮电大学 | Virtual mine risk-taking behavior implementation system and method based on multiple agents |
| WO2017176944A1 (en) * | 2016-04-05 | 2017-10-12 | Fractal Industries, Inc. | System for fully integrated capture, and analysis of business information resulting in predictive decision making and simulation |
| CN106777528A (en) * | 2016-11-25 | 2017-05-31 | 山东蓝光软件有限公司 | The holographic forecast method of mine air-required volume |
| CN108133289A (en) * | 2017-12-21 | 2018-06-08 | 中国铁建电气化局集团有限公司 | Tunnel ventilation control method and system based on environmental forecasting |
| US20200348038A1 (en) * | 2019-07-12 | 2020-11-05 | Johnson Controls Technology Company | Hvac system design and operational tool for building infection control |
| CN110633540A (en) * | 2019-09-24 | 2019-12-31 | 青岛理工大学 | Metal mine ventilation three-dimensional visual simulation aided decision control system and method |
| CN112684694A (en) * | 2020-11-15 | 2021-04-20 | 杭州哲达科技股份有限公司 | Blast furnace blast and TRT real-time monitoring and simulation control system based on digital twin body |
| CN112761700A (en) * | 2021-01-14 | 2021-05-07 | 成都茹附敬科技有限公司 | Training method of neural network for wind gear control of underground wind adjusting device |
| CN113356916A (en) * | 2021-07-08 | 2021-09-07 | 长安大学 | Mine air flow regulation and control virtual system based on digital twin technology and intelligent regulation and control method |
| CN113530589A (en) * | 2021-07-29 | 2021-10-22 | 西安重装韩城煤矿机械有限公司 | Intelligent local ventilation system and method for supplying air to coal mine driving face according to needs |
| CN116576527A (en) * | 2023-05-19 | 2023-08-11 | 国核电力规划设计研究院有限公司 | Ventilation method, ventilation device, electronic equipment and storage medium |
Non-Patent Citations (4)
| Title |
|---|
| 张楠;黄俊;全太锋;: "煤矿安全监管系统数据库设计与分析", 数字通信, vol. 1, no. 05, pages 134 - 135 * |
| 朱斌等: "综掘面风流智能调控数字孪生系统", 计算机集成制造系统, vol. 29, no. 6, pages 2006 - 2018 * |
| 陈龙等: "平行矿山:从数字孪生到矿山智能", 自动化学报, vol. 47, no. 7, pages 1633 - 1645 * |
| 龚晓燕: "数字孪生驱动的掘进工作面出风口风流智能调控系统", 煤炭学报, vol. 46, no. 4, pages 1331 - 1340 * |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114484822A (en) * | 2022-02-10 | 2022-05-13 | 中国海洋大学 | Ocean platform ventilation system control method based on temperature and hydrogen sulfide concentration control |
| CN114738031A (en) * | 2022-04-15 | 2022-07-12 | 北京科技大学 | Natural and mechanical ventilation coupled construction tunnel digital twin intelligent ventilation method |
| RU2798530C1 (en) * | 2022-05-04 | 2023-06-23 | федеральное государственное автономное образовательное учреждение высшего образования "Пермский национальный исследовательский политехнический университет" | System for controlling processes in an underground mining enterprise based on demand for electricity |
| CN115027641A (en) * | 2022-06-10 | 2022-09-09 | 江南造船(集团)有限责任公司 | Ship ventilation management method and system based on virtual-real interaction |
| CN115184563B (en) * | 2022-09-08 | 2022-12-02 | 北京中环高科环境治理有限公司 | Chemical workshop field data acquisition method based on digital twinning |
| CN115184563A (en) * | 2022-09-08 | 2022-10-14 | 北京中环高科环境治理有限公司 | Chemical workshop field data acquisition method based on digital twinning |
| CN115270532A (en) * | 2022-09-29 | 2022-11-01 | 山东新矿信息技术有限公司 | Data processing method and system based on mine intellectualization |
| CN115270532B (en) * | 2022-09-29 | 2022-12-23 | 山东新矿信息技术有限公司 | Data processing method and system based on mine intellectualization |
| CN115358494A (en) * | 2022-10-20 | 2022-11-18 | 中电建铁路建设投资集团有限公司 | Danger early warning method for subway shield underpass construction |
| CN115587542A (en) * | 2022-11-02 | 2023-01-10 | 四川奥凸水处理系统工程有限公司 | Underground water inversion simulation method, system, equipment and medium based on reinforcement learning |
| CN115788591A (en) * | 2023-02-08 | 2023-03-14 | 山东大学 | Monitoring and intelligent regulation and control system for dust and harmful gas in deep-buried long tunnel |
| CN116090076A (en) * | 2023-03-07 | 2023-05-09 | 四川省公路规划勘察设计研究院有限公司 | Gabion abutment building system under complex environment and rapid building method thereof |
| CN117411383A (en) * | 2023-12-15 | 2024-01-16 | 华夏天信智能物联股份有限公司 | Integrated mining variable-frequency speed regulation equipment control method and device and electronic equipment |
| CN117411383B (en) * | 2023-12-15 | 2024-02-20 | 华夏天信智能物联股份有限公司 | Integrated mining variable-frequency speed regulation equipment control method and device and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114000907A (en) | Mine ventilation equipment intelligent regulation and control system based on digital twin technology | |
| CN105082180B (en) | A kind of Tunnel testing robot and detection method | |
| CN111798051A (en) | Air quality space-time prediction method based on long-short term memory neural network | |
| CN108267172A (en) | Mining intelligent robot inspection system | |
| CN114673558B (en) | Coal mine driving face risk identification and intelligent pre-control system and method | |
| CN111079978B (en) | Coal and gas outburst prediction method based on logistic regression and reinforcement learning | |
| CN115570228B (en) | Intelligent feedback control method and system for welding pipeline gas supply | |
| CN115963763A (en) | Mine intelligent ventilation regulation and control system based on digital twin and data driving | |
| CN115453945A (en) | Comprehensive digging face intelligent ventilation control system based on mobile environment detection device | |
| CN112127934B (en) | Variable-frequency energy-saving ventilation system and method for high-altitude high-temperature extra-long tunnel construction | |
| Cheng et al. | Intelligent localization of a high-speed train using LSSVM and the online sparse optimization approach | |
| US9708899B2 (en) | Optimization of a multi-period model for valuation applied to flow control valves | |
| Zhang et al. | Optimization of backpropagation neural network under the adaptive genetic algorithm | |
| CN112650272B (en) | Coal mine underground unmanned aerial vehicle inspection information sensing method and sensing system based on 5G | |
| CN106779071A (en) | A kind of neutral net adaptive speed regulation method for Mine Ventilator | |
| CN116167531A (en) | Photovoltaic power generation prediction method based on digital twin | |
| CN113156492B (en) | Real-time intelligent early warning method applied to TBM tunnel rockburst disasters | |
| CN111273562A (en) | Method for realizing optimization of underground robot path based on membrane calculation | |
| CN113605866B (en) | Dynamic regulation and control system and method for mine gas extraction | |
| CN114742987A (en) | Automatic positioning control method and system for cutting of non-metallic materials | |
| CN113960925A (en) | Building energy consumption control method and device based on artificial intelligence | |
| CN117350103A (en) | Drilling parameter optimization method and system based on digital twin system | |
| CN112149255A (en) | Coal face hydraulic support working resistance prediction method based on LSTM neural network | |
| CN114973443B (en) | Inspection robot-based complex gas environment inspection system and method thereof | |
| Tariq et al. | Deep-AI soft sensor for sustainable health risk monitoring and control of fine particulate matter at sensor devoid underground spaces: A zero-shot transfer learning approach |
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 |