CN113356916B - Mine air flow regulation and control virtual system based on digital twin technology and intelligent regulation and control method - Google Patents
Mine air flow regulation and control virtual system based on digital twin technology and intelligent regulation and control method Download PDFInfo
- Publication number
- CN113356916B CN113356916B CN202110773534.8A CN202110773534A CN113356916B CN 113356916 B CN113356916 B CN 113356916B CN 202110773534 A CN202110773534 A CN 202110773534A CN 113356916 B CN113356916 B CN 113356916B
- Authority
- CN
- China
- Prior art keywords
- regulation
- data
- control
- wind
- intelligent
- 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.)
- Active
Links
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
-
- 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
- E21F1/02—Test models
Abstract
A mine wind flow regulation virtual system based on a digital twinning technology and an intelligent regulation method comprise a wind flow regulation system physical entity, a wind flow regulation system virtual entity, a wind flow intelligent decision system and twinning data, wherein the wind flow regulation system physical entity is used for restoring a scene of a coal mine excavation process, the wind flow regulation system virtual entity is used for optimizing and virtually debugging a digital twinning system, environment simulation, data flow simulation, sensor simulation and virtual debugging functions are provided, a learning evolution environment is provided for the digital twinning system, the wind flow intelligent decision system is responsible for processing the twinning data, and an intelligent regulation scheme is provided for a wind flow intelligent regulation device; the twin data are used for storing gas dust concentration and wind speed data acquired by the sensor in real time, current operating parameters of the wind drum and wind drum regulation and control parameters acquired by the wind flow intelligent decision system. The invention realizes the intelligent regulation and control of the wind flow of the fully-mechanized excavation face of the coal mine and provides an effective solution for optimizing the migration distribution of the wind flow field of the fully-mechanized excavation face.
Description
Technical Field
The invention relates to the technical field of mine wind current regulation, in particular to a mine wind current regulation virtual system and an intelligent regulation method based on a digital twinning technology.
Background
Coal occupies a leading position in the energy structure of China, and safe and efficient mining is the direction of high-quality development in the coal field. In the coal mining process, a large amount of gas and dust can be accumulated on the fully mechanized coal mining face, and how to reasonably and effectively reduce the concentration of the gas and the dust in a roadway and control the concentration of the gas and the dust in a safety range is one of main problems in the coal mining process. At present, mine operation adopts local fan to blow in gas and dust in the fresh air current's the mode more and dilutes the tunnel, and this kind of extensive formula "ventilation total amount" control mode ventilation inefficiency, dryer air outlet air current state can not be along with the real-time dynamic adjustment of tunnel operational environment, is difficult to guarantee the rational distribution in wind field to cause to combine to dig a gas and dust and gather, and then lead to gas, dust explosion, seriously threaten operation personnel's life safety. Therefore, the method for intelligently regulating and controlling the wind flow of the fully-mechanized excavation face of the coal mine has important practical significance for avoiding underground property loss and casualties and fundamentally solving the safety problem of the coal mine.
Disclosure of Invention
In order to overcome the technical problems, the invention aims to provide a mine wind flow regulation and control virtual system and an intelligent regulation and control method based on a digital twin technology, so that the intelligent regulation and control of the wind flow of the fully-mechanized excavation face of a coal mine are realized, and an effective solution is provided for optimizing the migration distribution of the wind flow field of the fully-mechanized excavation face.
In order to achieve the purpose, the invention adopts the technical scheme that:
a mine wind flow regulation and control virtual system based on a digital twin technology comprises a wind flow regulation and control system physical entity, a wind flow regulation and control system virtual entity, a wind flow intelligent decision system and twin data, wherein:
the wind flow regulation and control system physical entity is used for restoring a scene of a coal mine excavation process and comprises physical equipment, a wind flow regulation and control intelligent device and a plurality of sensors in the fully mechanized excavation process, wherein the physical equipment is provided with a communication interface and is connected with the sensors through the communication interface;
the intelligent air flow regulation and control device and the sensors can realize regulation and control of the front and back positions of the air outlet of the air duct away from the tunneling end surface, the angle deflection of the air outlet and the opening and closing of the caliber through PLC control;
the plurality of sensors comprise an air speed sensor, a gas and dust concentration sensor and a distance measuring sensor, the data acquisition instrument transmits data measured by various sensors to an upper computer monitoring and displaying system, so that the sensing and real-time acquisition of the data of the air speed, gas and dust concentration in different areas of the tunneling roadway and the on-line monitoring of the regulation and control state of the intelligent air flow regulation and control device are realized;
the virtual entity of the wind flow regulating and controlling system is used as an optimization and virtual debugging environment of the digital twin system, provides functions of environment simulation, data flow simulation, sensor simulation and virtual debugging, provides a learning evolution environment for the digital twin system, and is connected with the physical entity through data of each sensor; the virtual entity and the physical entity collect real-time status data of the physical entity based on an Open Platform Communication unified framework (opuca) protocol of a Programmable Logic Controller (PLC), and establish a mapping relationship with simulation data formed by the physical entity by calling a corresponding model in the data-driven virtual entity. Meanwhile, the control data of the intelligent air duct regulation and control device is transmitted by adopting an OPCUA protocol, and the movement, deflection and aperture opening and closing of the air duct are jointly controlled. By continuously iterating and optimizing the acquisition control process, the connection and dynamic interaction of real-time data of the physical entity and the virtual entity can be realized;
the wind flow intelligent decision system is responsible for processing twin data, generating decision information through data information, providing an intelligent regulation and control scheme for the wind flow intelligent regulation and control device, and obtaining iterative optimization of wind cylinder regulation and control parameters;
the twin data are used for storing gas dust concentration and wind speed data acquired by the sensor in real time, current operating parameters of the air duct and optimized air duct regulation and control parameters acquired by the air flow intelligent decision system through prediction or regulation and control rules.
The wind speed sensors respectively intercept 6 sections which are 5m, 7.5m, 10m, 15m, 20m and 25m away from the tunneling end face, 3 wind speed sensor nodes are arranged in each section, the height of a worker breathing zone is 1.5m away from the bottom of a roadway, the distance of the worker breathing zone is 1m away from the left wall of the roadway, the distance of the worker breathing zone is 1m away from the right wall of the roadway, and the distance of the worker breathing zone is 3m away from the right wall of the roadway, wherein 18 wind speed sensor nodes are arranged in total to comprehensively sense the wind speed state in the roadway; 1 wind speed sensor node is respectively arranged at the position with the height of 1.8m of a breathing zone at the position of a driver and at the air outlet of the wind flow dynamic regulation and control device to monitor the wind speed at the position of the driver and the air outlet in an important way, 2 sensors are arranged at the position, and 20 wind speed sensors are arranged in a roadway in total;
the gas sensor respectively intercepts 3 sections 5m, 7.5m and 10m away from the tunneling end face, and 3 gas sensor nodes are arranged in each section and are respectively at the height of 0.3m away from the bottom of the roadway; gas sensor nodes are arranged on the section 2m away from the tunneling end face, and are respectively 0.3m away from the left wall and 0.3m away from the right wall at the height of 0.3m away from the top plate; the height of the place 0.3m away from the bottom plate is one place 0.3m away from the right wall, and the total number of the gas sensors is 14;
the dust sensor respectively intercepts 6 sections 5m, 7.5m, 10m, 15m, 20m and 25m away from the tunneling end face, 3 dust sensor nodes are arranged in each section, namely, a position 1m away from the left wall of the roadway, a position 1m away from the right wall of the roadway and a position 3m away from the right wall of the roadway are respectively arranged at a position 1.5m away from the bottom of the roadway, and 18 dust sensor nodes are arranged in total to comprehensively sense the dust concentration distribution condition in the roadway; and 1 sensor is arranged at the position of a driver to sense the dust concentration at the position of the driver, and 19 dust sensors are arranged in total.
The intelligent wind flow decision system adopts an LSTM network to mine wind flow regulation twin data, predicts the change trend of GAs and dust concentration of a mine tunnel at the future time, regulates and optimizes regulation parameters in real time, sends early warning information if the GAs and dust concentration at the future time has an ascending trend, adopts a GA-BP neural network to generate a regulation decision scheme to analyze and optimize the regulation parameters of the air outlet of the air duct, simulates and verifies the optimization result, and adjusts the regulation parameters of the air outlet of the air duct in real time to adapt to the change of a mine operation environment.
The air flow regulation and control system collects the gas and dust concentration in the roadway and various regulation and control parameters of the air outlet of the air duct through sensors in the physical entity part, and converts the regulation and control parameters into standard twin data which can be used by a virtual entity and an air flow intelligent decision system after A/D conversion and data fusion processing; the virtual entity reads the concentration data of the processed gas and dust and the regulation and control parameters of the air outlet of the air duct by using twin data, and can visually simulate the real concentration distribution of the gas and dust in a physical tunnel by using particles with different sizes and colors in a virtual system by using the concentration data; the motion state of the physical wind flow regulating and controlling device can be simulated in real time by utilizing the regulating and controlling parameter data; and the intelligent wind flow decision system reads the processed gas and dust concentration data from the twin data and utilizes an early warning mechanism to perform safety judgment on the data. If the concentration exceeds the standard, sending early warning information in time, calling the existing regulation and control rule of the system background according to the early warning information, and generating a regulation and control scheme of the air outlet of the air drum; and if the concentration does not exceed the standard, predicting the GAs dust concentration at the future moment by using the LSTM network, and generating a reasonable decision scheme of the air outlet of the air duct by using a GA-BP neural network according to the prediction result. The intelligent service system sends out a regulation and control instruction in real time according to the decision scheme, and sends the regulation and control instruction to the air duct regulation and control PLC of the physical test system, so that the intelligent regulation and control of the air outlet parameters are realized, and a working cycle of air flow regulation and control is completed.
A regulation and control method of a mine air flow regulation and control virtual system based on a digital twin technology comprises the following steps;
step S1, acquiring the regulation and control parameters and environmental parameter data of the air duct, preprocessing the data and determining the topological structure of the neural network;
s2, encoding the weight and the threshold of the neural network obtained in the S1 to obtain an initial population;
step S3, encoding the initial value of step S2, introducing a genetic algorithm and calculating fitness, if the termination condition is met, decoding to obtain an optimal weight and a threshold;
step S4, training the neural network by using the optimal weight and the threshold obtained in the step S3 to obtain an output layer result, and calculating the output error of the neural network;
step S5, if the output error meets the termination condition, outputting an output layer;
and step S6, if the output error does not meet the termination condition, regenerating the output layer result according to the weight value and the threshold value updated by the optimizer until the output error meets the termination condition.
The step of acquiring the data and preprocessing the data comprises the following steps: dividing the data into training set data and test set data, and preprocessing the training set data and the test set data through a normalization formula and an inverse normalization formula; wherein the normalization formula is:
in the formula, x i ' is data after standard pyramidal; x is the number of i Is original data; x is the number of min Is the minimum value of the sample data; x is the number of max Is the maximum value of the sample data.
The inverse normalization formula is:
in the formula, x i ' is normalized data; x is the number of i Is original data; mu is the mean value of sample data; σ is the standard deviation of the sample data.
A step of determining the topology of the neural network in the step S1; constructing a BP neural network and constructing hidden layer nodes; defining a network structure as a three-layer BP neural network structure with a single hidden layer, wherein 5 input layer nodes are respectively the gas concentration in a dead angle area of a tunneling end face, the dust concentration at a driver, the dust concentration at an air return side, the air speed at the air return side and the air speed at the driver; the number of the output layer nodes is 4, and the output layer nodes are respectively the distance from the air outlet of the air duct to the tunneling end face, the caliber of the air outlet of the air duct, the right deflection angle of the air outlet and the upward deflection angle of the air outlet; the hidden layer nodes are determined by adopting a method of combining a trial-and-error method and an empirical method, the optimal number of the hidden layer nodes is estimated by using an empirical formula, and then the optimal value of the final hidden layer node number is determined by comparing the mean square errors of the network under different node numbers by using the trial-and-error method on the basis. Wherein the empirical formula is
In the formula, k is the number of hidden layer nodes; n is the number of nodes of the input layer; m is the number of output layer nodes; alpha is a constant between 1 and 10.
The encoding mode of the genetic algorithm in the step S3 adopts floating point number encoding, and the chromosome length is 134 bits. Wherein, the first 65 bits are the weight codes from the input layer to the hidden layer; 66-78 bits are hidden layer threshold codes; 79-130 bits are weight codes from a hidden layer to an output layer; 131-134 bits are output layer threshold codes.
In the step S3, the fitness calculation of the genetic algorithm converts the reciprocal of the network training mean square error into a fitness function value, that is, the fitness function of each chromosome is:
wherein F represents a fitness value; n represents the neural network training sample capacity; t represents the true value of the sample; p denotes a sample prediction value.
The selection operator adopts a binary tournament selection algorithm, only the fitness values of the two selected individuals need to be sequenced, the crossover operator adopts circular crossover, and compared with other crossover operators, the crossover operator only needs to operate one chromosome in the crossover process to obtain the two chromosomes, so that the crossover process is simple, the algorithm complexity can be reduced, and the algorithm solving efficiency can be improved; the mutation operator adopts a non-uniform mutation operator, random disturbance can be performed on values of all gene positions of the chromosome once during mutation operation, the disturbed result is taken as a new value on the gene position to form a new chromosome, and compared with other mutation operators, the mutation operator can perform local search on a key area where the optimal solution exists.
The cross probability P c And the probability of variation P m With adaptive adjustment of P c And P m The method of (1) is represented by the following formula:
in the above formula, F max The maximum value of fitness in the population; f m Is the greater fitness value in the two chromosomes involved in the crossover; f avg The average value of fitness in the population; f is the fitness value of participating in the variant chromosome; f min Is the minimum value of fitness in the population.
In step S3, the genetic algorithm is stopped when the condition setting iteration is performed 300 times.
The activation function of the BP neural network part in step S4 is a Sigmoid activation function, and the function form is as follows:
the loss function of the neural network adopts a mean square error as the loss function, and the expression of the mean square error is as follows:
where MSE represents the mean square error; n represents a sample capacity; t represents the true value of the sample; p denotes a sample prediction value.
The optimizer of the neural network in the step S6 adopts Adam algorithm.
The invention has the beneficial effects.
According to the invention, the interaction and the co-fusion of a physical system and a virtual system are realized by constructing a digital twin model of a coal mine fully-mechanized excavation face local ventilation system, and the severe and complex operation environment of a coal mine roadway is represented. In order to enable the digital twin model to have the capabilities of autonomous learning, autonomous prediction and autonomous decision, the intelligent regulation and control of the coal mine fully-mechanized excavation face wind flow are realized by utilizing real-time data and historical data of a sensor, adopting key technologies such as regulation and control rule mining based on big data, prediction of GAs dust concentration based on a deep learning Long-Short Term Memory network (LSTM), air drum air outlet parameter decision based on a genetic-back propagation network (GA-BP) and the like, and by means of real-time simulation, virtual-real mapping, data fusion, iterative optimization and the like, an effective solution way is provided for optimizing migration distribution of the fully-mechanized excavation face wind flow field.
Description of the drawings:
fig. 1 is a general structural diagram of a digital twin mine wind flow regulation virtual system according to the present invention.
FIG. 2 is a diagram of a sensor node location arrangement of the present invention.
Fig. 3 is an operation flow of the intelligent wind flow decision system of the present invention.
Fig. 4 shows the movement parameters of the virtual entity of the wind flow regulating device.
Fig. 5 is a caliber opening and closing mechanism of the wind flow regulating device.
Fig. 6 is a simplified model of the aperture opening and closing pull rod mechanism.
FIG. 7 is a twin data information fusion model of the wind flow regulation system.
Fig. 8 is a technical route for obtaining the wind flow regulation rule.
FIG. 9 is a flow chart of GA-BP neural network decision-making regulation algorithm.
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 only a part of the embodiments of the present invention, and not all of the embodiments. 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.
The embodiment of the invention is as follows, see fig. 1, and comprises:
step 1: and building a physical entity.
The physical entity comprises physical equipment, an intelligent wind flow regulating and controlling device and a plurality of sensors in the fully mechanized excavating process, wherein the physical equipment is provided with a communication interface and is connected with the sensors through the communication interface; the specific position of the sensor is shown in FIG. 2; 1-6 sections, wherein three sensors of the same type are respectively arranged on each section, and the height of each sensor is 1.5m of the height of the pedestrian respiration belt; and 7# -11# measuring points are respectively provided with one sensor of the same type. The distance between the No. 1 section and the tunneling end face is 5m, the distance between the No. 2 section and the No. 1 section is 2.5m, the distance between the No. 3 section and the No. 2 section is 2.5m, the distance between the No. 4 section and the No. 3 section is 5m, the distance between the No. 5 section and the No. 4 section is 5m, and the distance between the No. 6 section and the No. 5 section is 5 m. And the 7# measuring point is positioned at the air outlet of the air flow state regulating and controlling device. And the 8# measuring point is positioned at the driver position of the heading machine. And the measuring points 9#, 10#, and 11# are respectively positioned at the upper left corner, the upper right corner and the lower right corner of the tunneling section.
Step 2: the specific implementation of the virtual entity of the wind flow control system comprises the following steps:
step (1): establishing a three-dimensional model consistent with a physical entity by using 3dsMax three-dimensional modeling software;
step (2): rendering each part by using a rendering function of 3dsMax to ensure that the color and the material are consistent with those of a physical entity;
and (3): motion constraint is added to the three-dimensional model, and motion animation is made by opening and closing the air duct aperture of the air flow regulating and controlling device. Fig. 4 is a corresponding relationship between the motion parameters of the wind flow control device and the object controlled by the virtual wind flow control device.
The aperture opening and closing mechanism of a single blade of the air flow regulating device is shown in fig. 5, and the mechanism takes a motor as an input piece and realizes the opening and closing movement of the aperture of the air duct through a bevel gear mechanism, a lead screw nut mechanism and a pull rod mechanism. A simplified model of a single blade caliber opening and closing pull rod mechanism of the wind flow control device is shown in figure 6, the mechanism consists of blades, connecting rods, sliding blocks and a lead screw, A, B, C are hinge central points of the blades and a rotating ring, the blades and the connecting rods, and the sliding blocks and the connecting rods respectively, w is the length of a half blade, r is the radius of the rotating ring, s is the displacement of the sliding blocks, and theta is the displacement of the sliding blocks 1 Is the angle of the blade to the horizontal, theta 2 Is the included angle between the connecting rod and the horizontal direction.
The opening and closing pull rod mechanism of the air outlet cylinder caliber can be provided with an OABCO closed ring, and then has a vector equation
The plural form is:
the X-axis and Y-axis scalar relation obtained by Euler transformation of the above formula is as follows:
the relationship between the aperture R of the air duct and the rotation angle of the blade in the horizontal direction is as follows:
R＝r+2wsin(θ 1 )
the simplified expression of the nonlinear transcendental equation system of the air duct caliber and the slide block displacement is as follows:
and (3) applying Taylor formula series expansion and omitting high-order terms, wherein the implicit function expression of the air duct aperture and the slider displacement is obtained as follows:
8wl(w+l)-l(R-r) 2 -w(R+r) 2 -s＝0
the corresponding relation can be established between the slide block displacement and the air duct aperture according to the formula, so that the mutual solution can be realized, and further the synchronous display of the virtual model air duct aperture data can be realized.
And 4, step 4: importing the three-dimensional model into Unity 3D;
and 5: building a scene, lights and cameras in Unity 3D;
step 6: writing a control program for horizontal and vertical rotation of the wind flow regulating and controlling device by using the Visual Studio 2017;
and 7: an Animation module in the Unity3D establishes an Animation hierarchical relationship of opening and closing of the wind tube aperture, and controls the opening and closing movement of the wind tube through an Animation frame. The animation with sequence frames has s frames, the animation is played from a certain time t0 (the s0 th frame) to the time t, the playing is stopped, the playing speed is v, and then the animation playing frames at the time t0+ Δ t are:
the minimum diameter of the wind tube is D min Maximum diameter of D max Air duct initial movement caliber D at a certain time 0 The corresponding animation frame is
The animation frame corresponding to the target moving aperture D is
The diameter of the wind tube is D 0 Changing to D is an animation frame from
Frame playing to s D A frame;
and 8: adding Particle simulation of gas and dust in a roadway by using a Particle System module, setting the generation speed and the life cycle of the gas and the dust, and setting the gas to float upwards and the dust to fall downwards;
and step 9: the Visual Studio2017 is used to write a control program for the movement, agglomeration and dispersion phenomena of the particles and the collision of the particles.
When gas or dust moves in a roadway environment, three types of collision exist, namely particle-particle collision, particle-roadway collision and particle-ground collision. These all need to be set up by means of the Collision module of the Unity3D particle system, particle adding rigid Collider, lane adding grid Collider (Mesh Collider), which can trigger different actions when the Collision occurs between particles, between particles and lanes or between particles and the ground, as shown in the following table.
The gas or dust concentration simulation particle system initializes some necessary attributes, such as quantity, size, color, transparency, life time and the like, and the sensing data of the physical world is visually displayed in a virtual space in real time through the combined calculation of the particle attributes.
In the gas and dust concentration simulation particle system, the particles can be divided into two parts of old particles and new particles. The old particles are the particles left after the concentration simulation particles collected by the sensor before in the process of simulating the concentration of the gas and the dust change along with the time and the life cycle disappears. The new particles are the particles needed by the area sensor to collect concentration simulation at the current moment.
The newly generated particle number NewNum at any time (the time represents the sampling period of the sensor) of the particle system is the particle number newly generated required for simulating the acquisition concentration of the current sensor minus the particle number LiveNum left after the extinction of the concentration simulation particles at the last time:
NewNum(t)＝NeedNum(t)-LiveNum(t-1)
in order to improve the user experience and enhance the visual effect of the simulation of the gas and dust concentration particles, the size of the particles in the particle system is variable within the maximum and minimum size parameter ranges, because the phenomena of conglomeration or dispersion and the like can occur in the actual movement process regardless of gas or dust particles, and the phenomenon can be simulated by the change of the particle size. The random initialization particle size formula is as follows:
StartSize＝MinSize+Rand(0，1)*(MaxSize-MinSize)
the particle agglomeration and dispersion are simulated in the particle movement process, the current time of the particle is a random value in a variation range, and the calculation formula is as follows:
wherein NewSize represents the particle size at the new moment, StartSize represents the initial particle size, MinSize and MaxSize are respectively the maximum and minimum particle diameters, Rand (0,1) is a random number between 0 and 1, k is a quantized quantity, and the value is 0 or 1, and respectively represents particle agglomeration and dispersion.
The size of the particles in the particle system at a certain time depends on whether it is in the agglomeration process, in which the current particle size is a random value in the upper limit of the initialized particle size and the particle size range, i.e. the size range is [ StartSize, MaxSize ], or in the dispersion process, in which the current particle size is a random value in the lower limit of the particle size range and the initialized particle size, i.e. the size range is [ MinSize, StartSize ].
For gas, such as colorless and odorless gas, it is difficult to simulate the gas by using a particle system, and for convenient observation, it is necessary to change the color attributes of the particles to represent the gas concentration condition in the region. In order to obtain good gas visual simulation effect, the particle color should be randomly changed, and the particle initialization color (the particle color is expressed by using three primary colors of RGB and transparency) is:
InitialColor(R,G,B)＝Rand(0，1)*VarColor(R,G,B)
InitialColor(Alpha)＝1.0
wherein Alpha is the transparency of the particles, and VarColor (R, G, B) is the variation range of the three primary colors of the particle colors.
Step 10: writing the import and the output of data by using Visual Studio 2017;
step 11: and creating a human-computer interaction interface.
(2) The specific implementation of the air duct air outlet parameter regulation and control rule obtaining algorithm comprises the following steps:
the invention adopts finite element flow field analysis technology and rough set theory, takes parameters of different tunnel sizes, air outlet deflection, caliber size, moving distance and the like as condition attributes, obtains the gas and dust concentration through numerical simulation calculation, and takes the gas and dust concentration as decision attributes, thereby establishing a decision information system, wherein the technical route is as shown in figure 8, and the decision information system is divided into three stages:
1) according to the actual ventilation condition and the fluid mechanics theory of the comprehensive data surface, finite element software is adopted to establish finite element models of the air flow, the gas and the dust field under different working conditions, the numerical simulation calculation of the air flow, the gas and the dust field is carried out, and the feasibility and the effectiveness of the finite element models are verified through underground actual measurement;
2) carrying out numerical simulation analysis on a gas dust field under the comprehensive change of the aperture of an air outlet of the air duct, the horizontal deflection angle and the vertical deflection angle in different cutting modes to obtain sample data, and preprocessing the data to establish a decision information system;
3) and taking the aperture of an air outlet of the air duct, the horizontal deflection angle and the vertical deflection angle as conditional attributes, taking the air speed gas dust concentration as a decision attribute, discretizing the sample data, and extracting the air flow regulation and control rules of the fully-mechanized excavation face under different working conditions by using an air flow regulation and control rule acquisition algorithm based on particle calculation.
Setting the comprehensive digging surface wind flow regulation and control decision information system as S ═ U, A, V and f,
and G 2 Two knowledge items are (ψ, m (ψ)),
to represent
Number of middle objects, granule G 1 、G 2 The corresponding decision rule is
Defining rule coverage and confidence respectively as follows:
the algorithm is input into a comprehensive digging surface wind flow regulation and control decision information table S ═ U, C ═ D, V, f, wherein C represents a condition attribute set and the caliber of an air outlet of an air duct, a horizontal deflection angle and a vertical deflection angle, and D represents a decision attribute, the height of a breathing zone of a person on the air return side in the process under each cutting stage and the concentration of gas and dust at a driver; the output of the algorithm is a simplest decision rule for regulating and controlling the wind flow. The algorithm comprises the following specific steps:
step 1: calculating a basic particle library GK corresponding to the condition attribute set and a basic particle library GD corresponding to the decision attribute set D, and enabling M to be phi, GA to be phi, GR' to be phi and lambda to be 1;
step 2: if GD ═ φ, go to step 8;
and step 3: for the
Turning to the step 4 to the step 7;
and 4, step 4: obtaining a knowledge particle library GR;
and 5: for all granules in the knowledge granule library GR
If it is not
The simplest decision rule is output:
and all the objects covered by the rule are added into the set M, and GA is made to be GA and G r ，GR＝GR/G r (ii) a Otherwise, turning to step 6;
step 6: judging M and GR, if M ═ M (psi) i ) And GR is not equal to phi, then in the knowledge granule library GR, all the requirements are met
And is
Granule of (1)
The following operations are performed:
(a) let G ═G i ∧G j If particle, if
Then GR ' ═ GR '. goug ';
(b) if GR ' ≠ Φ, let GR be GR ', GR ' ═ Φ, λ + 1;
and 7: let GD be GD/GD i Turning to step 2 if M is equal to phi and GA is equal to phi;
and 8: the algorithm ends.
(3) The concrete implementation of the gas and dust concentration prediction in the invention comprises the following steps:
the method adopts the LSTM network to mine the potential time sequence relation of the gas dust concentration data, predicts the change trend of the gas dust concentration at the future moment, and provides a reasonable regulation and control scheme for the wind current regulation and control physical device. Therefore, historical monitoring data of gas dust concentration acquired by a physical testing system is used as input of an LSTM network, gas dust concentration at a future moment is used as output, a 3-layer network structure is adopted, namely, only one layer of hidden layer is adopted, and the activation function and the network optimizer are determined to be a tanh function and a batch gradient descent algorithm respectively by changing the activation function and the optimizer and comparing error magnitude of training results. When the gas dust concentration is predicted, firstly, input data is mapped into a multidimensional vector, then the multidimensional vector is input into the neuron units of the LSTM layers, the output value of the LSTM neuron unit of each layer serves as the input of the next LSTM neuron unit, and the output of the last LSTM layer is mapped into a two-dimensional vector through the full-connection layer, namely the gas dust concentration at the future moment. The LSTM network algorithm is implemented as follows:
step 1: forget door f for calculating t moment t Comprises the following steps:
f t ＝σ(W f ·[h t-1 ,x t ]+b f )
where σ is the forgetting gate activation function, W f To forget the gate weight matrix, [ h t-1 ,x t ]To forget to input a value, b f Biasing the item for the forgetting gate;
and 2, step: input gate i for calculating time t t Comprises the following steps:
i t ＝σ(W i ·[h t-1 ,x t ]+b i )
wherein, W i As a weight matrix of output gates, b i Biasing the item for the forgetting gate;
and step 3: output gate o for calculating time t t Comprises the following steps:
o t ＝σ(W o ·[h t-1 ,x t ]+b o )
wherein, W o As a weight matrix of output gates, b o Biasing the item for the forgetting gate;
and 4, step 4: computing t Temporal long-term memory state c t Comprises the following steps:
c t ＝f t ·c t-1 +i t ·c t
wherein, c t-1 The long-term memory state at the last moment;
and 5: finally obtaining concentration values h of gas and dust at t moment t Comprises the following steps:
h t ＝o t ·tanh(c t )
(4) the specific implementation of the decision regulation and control of the air duct air outlet parameter based on the GA-BP neural network comprises the following steps:
according to the invention, the strong fitting capability of the BP neural network is utilized to mine the intrinsic relation between the gas dust concentration and the air outlet parameter of the air duct, and the air outlet parameter of the air duct is autonomously decided, so that the intelligent regulation and control of the gas and dust concentration are realized, and the occurrence of potential safety hazards is prevented. Considering that when the BP neural network makes a decision on the parameters of the air outlet of the air duct, the initial weight and the threshold of the network are random numbers between [ -1,1], and the initial weight and the threshold have a great influence on the learning precision of the network, the BP neural network is optimized by adopting a genetic algorithm, and the optimal initial weight and the threshold of the BP neural network are obtained by utilizing the strong searching capability of the BP neural network, so that the training time of a model is shortened, and the precision of the decision on the parameters of the air outlet of the air duct is improved. The flow chart of the GA-BP neural network decision control algorithm is shown in FIG. 9.
When the GA-BP neural network is designed, influence relations between GAs dust concentration and air outlet parameters of the air duct are researched, the BP neural network input layer nodes are determined to be the air speed of pedestrians, the dead-angle area GAs concentration and the dust concentration of drivers in a coal mine tunnel, the output layer nodes are determined to be the distance from the air outlet of the air duct to a tunneling end, the caliber of an air outlet, a horizontal deflection angle and a vertical deflection angle, and the number of hidden layer nodes is determined by using an empirical formula. And comparing the error magnitude of the training result by changing the activation function and the optimizer, and determining that the activation function and the network optimizer are a sigmoid function and a variable momentum estimation algorithm respectively.
In addition, because the initial weight and the threshold of the BP neural network are floating point numbers and belong to the continuity problem, the genetic algorithm for optimizing the BP neural network adopts floating point number coding, the reciprocal of the mean square error of the training of the BP neural network is used as a fitness function, and a selection operator, a crossover operator and a mutation operator respectively select a binary competitive bidding match selection operator, a circular crossover operator and a non-uniform mutation operator. Meanwhile, in order to obtain better optimizing effect of the genetic algorithm, the invention adopts a cross probability P C And the mutation probability P m The adaptive method which can change along with the adaptability value is shown as the following formula:
wherein, F max The maximum value of the fitness value in the current population; f m Is the greater fitness value in the two chromosomes involved in the crossover; f avg The average value of the fitness value in the current population is obtained; f is the fitness value of participating in the variant chromosome; f min Is the minimum value of the fitness value in the current population. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art can still implement the present invention in other specific waysModifications or equivalents may be made thereto without departing from the spirit and scope of the invention, which is set forth in the claims appended hereto.
Claims (8)
1. A mine wind flow regulation virtual system based on a digital twin technology is characterized by comprising a physical entity of a wind flow regulation system, a virtual entity of the wind flow regulation system, an intelligent wind flow decision system and twin data, wherein:
the wind flow regulation and control system physical entity is used for restoring a scene of a coal mine excavation process and comprises physical equipment, a wind flow regulation and control intelligent device and a plurality of sensors in the fully mechanized excavation process, wherein the physical equipment is provided with a communication interface and is connected with the sensors through the communication interface;
the intelligent air flow regulation and control device and the sensors can realize regulation and control of the front and back positions of the air outlet of the air duct away from the tunneling end surface, the angle deflection of the air outlet and the opening and closing of the caliber through PLC control;
the plurality of sensors comprise an air speed sensor, a gas sensor, a dust sensor and a distance measuring sensor, the data acquisition instrument transmits data measured by various sensors to an upper computer monitoring display system, so that the sensing and real-time acquisition of the data of the air speed, gas and dust concentration in different areas of the tunneling roadway and the online monitoring of the regulation and control state of the intelligent air flow regulation and control device are realized;
the virtual entity of the wind flow regulating and controlling system is used as an optimization and virtual debugging environment of the digital twin system, provides functions of environment simulation, data flow simulation, sensor simulation and virtual debugging, provides a learning evolution environment for the digital twin system, and is connected with the physical entity through data of each sensor;
the wind flow intelligent decision system is responsible for processing twin data, generating decision information through data information, providing an intelligent regulation and control scheme for the wind flow intelligent regulation and control device, and obtaining iterative optimization of wind cylinder regulation and control parameters;
the twin data is used for storing gas dust concentration and wind speed data acquired by the sensor in real time, current operating parameters of the air duct and optimized air duct regulation and control parameters acquired by the air flow intelligent decision system through prediction or regulation and control rules;
the intelligent wind flow decision system adopts an LSTM network to mine wind flow regulation twin data, predicts the change trend of GAs and dust concentration of a mine tunnel at the future time, regulates and optimizes regulation and control parameters in real time, and sends early warning information if the GAs and dust concentration at the future time has an ascending trend;
the air flow regulation and control system collects the gas and dust concentration in the roadway and various regulation and control parameters of the air outlet of the air duct through sensors in the physical entity part, and converts the regulation and control parameters into standard twin data which can be used by a virtual entity and an air flow intelligent decision system after A/D conversion and data fusion processing; the virtual entity reads the concentration data of the processed gas and dust and the regulation and control parameters of the air outlet of the air duct by using twin data, and can visually simulate the real concentration distribution of the gas and dust in a physical tunnel by using particles with different sizes and colors in a virtual system by using the concentration data; the motion state of the physical wind current regulation and control device can be simulated in real time by utilizing the regulation and control parameter data; the wind flow intelligent decision system reads the processed gas and dust concentration data from the twin data and utilizes an early warning mechanism to perform safety judgment on the data; if the concentration exceeds the standard, early warning information is sent out in time, the existing regulation and control rule of the system background is called according to the early warning information, and a regulation and control scheme of the air outlet of the air duct is generated; if the concentration does not exceed the standard, predicting the GAs dust concentration at the future moment by using an LSTM network, generating a reasonable decision scheme of an air outlet of an air duct by using a GA-BP neural network according to a prediction result, sending a regulation and control instruction by an intelligent service system in real time according to the decision scheme, and sending the regulation and control instruction to an air duct regulation and control PLC controller of a physical test system to realize intelligent regulation and control of air outlet parameters, so that a working cycle of air flow regulation and control is completed;
the LSTM network algorithm implementation steps are as follows:
step 1: forget door f for calculating t moment t Comprises the following steps:
f t ＝σ(W f ·[h t-1 ,x t ]+b f )
where σ is the forgetting gate activation function, W f To forget the gate weight matrix, [ h t-1 ,x t ]To forget to input a value, b f Biasing the item for the forgetting gate;
step 2: input gate i for calculating time t t Comprises the following steps:
i t ＝σ(W i ·[h t-1 ,x t ]+b i )
wherein, W i As a weight matrix of output gates, b i Biasing the item for the forgetting gate;
and step 3: output gate o for calculating time t t Comprises the following steps:
o t ＝σ(W o ·[h t-1 ,x t ]+b o )
wherein, W o As a weight matrix of output gates, b o Biasing the item for the forgetting gate;
and 4, step 4: calculating the long-term memory state c at time t t Comprises the following steps:
c t ＝f t ·c t-1 +i t ·c t
wherein, c t-1 The long-term memory state at the last moment;
and 5: finally obtaining concentration values h of gas and dust at t moment t Comprises the following steps:
h t ＝o t ·tanh(c t )
the specific implementation of the air duct air outlet parameter decision regulation of the GA-BP neural network comprises the following steps:
the method comprises the steps of utilizing the strong fitting capacity of a BP neural network to dig the internal relation between the gas dust concentration and the air outlet parameter of the air drum, carrying out autonomous decision making on the air outlet parameter of the air drum, achieving intelligent regulation and control of the gas dust concentration, preventing the occurrence of potential safety hazards, enabling the initial weight and the threshold of the network to be random numbers between < -1 > and 1 >, optimizing the BP neural network by adopting a genetic algorithm, and obtaining the optimal initial weight and the threshold of the BP neural network.
2. The mine wind flow regulation virtual system based on the digital twinning technology is characterized in that 6 sections of 5m, 7.5m, 10m, 15m, 20m and 25m away from a tunneling end face are respectively intercepted by the wind speed sensors, 3 wind speed sensor nodes are arranged in each section, the height of a worker breathing zone is 1.5m away from the bottom of a roadway, the height of the worker breathing zone is 1m away from the left wall of the roadway, the height of the worker breathing zone is 1m away from the right wall of the roadway, the distance of the worker breathing zone is 3m away from the right wall of the roadway, and 18 wind speed sensor nodes are arranged in total to comprehensively sense the wind speed state in the roadway; 1 wind speed sensor node is respectively arranged at the position with the height of 1.8m of a breathing zone at the position of a driver and at the air outlet of the wind flow dynamic regulation and control device to monitor the wind speed at the position of the driver and the air outlet in an important way, 2 sensors are arranged at the position, and 20 wind speed sensors are arranged in a roadway in total;
the gas sensor respectively intercepts 3 sections 5m, 7.5m and 10m away from the tunneling end face, and 3 gas sensor nodes are arranged in each section and are respectively at the height of 0.3m away from the bottom of the roadway; gas sensor nodes are arranged on the section 2m away from the tunneling end face, and are respectively 0.3m away from the left wall and 0.3m away from the right wall at the height of 0.3m away from the top plate; the height of the place 0.3m away from the bottom plate is one place 0.3m away from the right wall, and the total number of the gas sensors is 14;
the dust sensor respectively intercepts 6 sections 5m, 7.5m, 10m, 15m, 20m and 25m away from the tunneling end face, 3 dust sensor nodes are arranged in each section, namely, a position 1m away from the left wall of the roadway, a position 1m away from the right wall of the roadway and a position 3m away from the right wall of the roadway are respectively arranged at a position 1.5m away from the bottom of the roadway, and 18 dust sensor nodes are arranged in total to comprehensively sense the dust concentration distribution condition in the roadway; and 1 sensor is arranged at the position of a driver to sense the dust concentration at the position of the driver, and 19 dust sensors are arranged in total.
3. The regulation and control method of the mine shaft air flow regulation and control virtual system based on the digital twin technology is characterized by comprising the following steps;
step S1, acquiring the regulation and control parameters and environmental parameter data of the air duct, preprocessing the data and determining the topological structure of the neural network;
step S2, encoding the weight and the threshold of the neural network obtained in the step S1 to obtain an initial population;
step S3, encoding the initial value of step S2, introducing a genetic algorithm and calculating fitness, if the termination condition is met, decoding to obtain an optimal weight and a threshold;
step S4, training the neural network by using the optimal weight and the threshold obtained in the step S3 to obtain an output layer result, and calculating the output error of the neural network;
step S5, if the output error meets the termination condition, outputting an output layer;
and step S6, if the output error does not meet the termination condition, regenerating the output layer result according to the weight value and the threshold value updated by the optimizer until the output error meets the termination condition.
4. The method for regulating and controlling the mine air flow regulation and control virtual system based on the digital twinning technology as claimed in claim 3, wherein the step of acquiring and preprocessing data comprises: dividing the data into training set data and test set data, and preprocessing the training set data and the test set data through a normalization formula and an inverse normalization formula; wherein the normalization formula is:
in the formula, x i ' is data after standard pyramidal; x is the number of i Is original data; x is the number of min Is the minimum value of the sample data; x is the number of max Is the maximum value of the sample data;
the inverse normalization formula is:
in the formula, x i ' is normalized data; x is the number of i Is original data; mu is the mean value of sample data; σ is the standard deviation of the sample data.
5. The method for regulating and controlling the mine shaft airflow regulation and control virtual system based on the digital twin technology as claimed in claim 3, wherein the step of determining the topological structure of the neural network in the step S1; constructing a BP neural network and constructing hidden layer nodes; defining a network structure as a three-layer BP neural network structure with a single hidden layer, wherein 5 input layer nodes are respectively the gas concentration in a dead angle area of a tunneling end face, the dust concentration at a driver, the dust concentration at an air return side, the air speed at the air return side and the air speed at the driver; the number of the output layer nodes is 4, and the output layer nodes are respectively the distance from the air outlet of the air duct to the tunneling end face, the caliber of the air outlet of the air duct, the right deflection angle of the air outlet and the upward deflection angle of the air outlet; the hidden layer nodes are determined by combining a trial-and-error method and an empirical method, the optimal number of the hidden layer nodes is estimated by using an empirical formula, then the trial-and-error method is used on the basis, the mean square error of the network under different node numbers is compared, and the optimal value of the number of the hidden layer nodes is determined finally, wherein the empirical formula is
Wherein k is the number of nodes of the hidden layer; n is the number of nodes of the input layer; m is the number of output layer nodes; alpha is a constant between 1 and 10.
6. The method for regulating and controlling the mine wind current regulation and control virtual system based on the digital twin technology according to claim 3, wherein the encoding mode of the genetic algorithm in the step S3 adopts floating point number encoding, the chromosome length is 134 bits, and the first 65 bits are weight encoding from an input layer to a hidden layer; 66-78 bits are hidden layer threshold codes; 79-130 bits are weight codes from a hidden layer to an output layer; 131-134 bits are output layer threshold codes;
in the step S3, the fitness calculation of the genetic algorithm converts the reciprocal of the network training mean square error into a fitness function value, that is, the fitness function of each chromosome is:
wherein F represents a fitness value; n represents the neural network training sample capacity; t represents the true value of the sample; p represents a sample prediction value;
the selection operator adopts a binary tournament selection operator, only the fitness values of the two selected individuals need to be sequenced, the crossover operator adopts circular crossover, and compared with other crossover operators, the crossover operator only needs to operate one chromosome in the crossover process to obtain two chromosomes, so that the crossover process is simple, the algorithm complexity can be reduced, and the algorithm solving efficiency can be improved; the mutation operator adopts a non-uniform mutation operator, random disturbance can be performed on values of all gene positions of the chromosome once during mutation operation, the disturbed result is taken as a new value on the gene position to form a new chromosome, and compared with other mutation operators, the mutation operator can perform local search on a key area where the optimal solution exists;
cross probability P c And the probability of variation P m With adaptive adjustment of P c And P m The method (1) is represented by the following formula:
in the above formula, F max The maximum value of fitness in the population; f m In two chromosomes participating in crossoverA larger fitness value; f avg The average value of fitness in the population; f is the fitness value of participating in the variant chromosome; f min Is the minimum value of fitness in the population;
in step S3, the genetic algorithm is stopped when the partial termination condition setting iteration is performed 300 times.
7. The method for regulating and controlling the mine airflow regulation and control virtual system based on the digital twin technology as claimed in claim 3, wherein the activation function of the BP neural network part in the step S4 is a Sigmoid activation function, and the function form is as follows:
the loss function of the neural network adopts a mean square error as the loss function, and the expression of the mean square error is as follows:
where MSE represents the mean square error; n represents a sample capacity; t represents the true value of the sample; p denotes a sample prediction value.
8. The method for regulating and controlling the mine air flow regulation and control virtual system based on the digital twin technology as claimed in claim 3, wherein the optimizer of the neural network in the step S6 adopts an Adam algorithm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110773534.8A CN113356916B (en) | 2021-07-08 | 2021-07-08 | Mine air flow regulation and control virtual system based on digital twin technology and intelligent regulation and control method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110773534.8A CN113356916B (en) | 2021-07-08 | 2021-07-08 | Mine air flow regulation and control virtual system based on digital twin technology and intelligent regulation and control method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113356916A CN113356916A (en) | 2021-09-07 |
| CN113356916B true CN113356916B (en) | 2022-09-16 |
Family
ID=77538557
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110773534.8A Active CN113356916B (en) | 2021-07-08 | 2021-07-08 | Mine air flow regulation and control virtual system based on digital twin technology and intelligent regulation and control method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113356916B (en) |
Families Citing this family (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113852204B (en) * | 2021-10-13 | 2022-06-07 | 北京智盟信通科技有限公司 | Transformer substation three-dimensional panoramic monitoring system and method based on digital twinning |
| CN114109465A (en) * | 2021-10-20 | 2022-03-01 | 中国计量大学 | Intelligent analysis decision-making method for mine ventilation system |
| CN113835460B (en) * | 2021-11-23 | 2022-05-03 | 中铁建设集团有限公司 | Data computer lab intelligence environmental control system based on hydrodynamics analog digital twin |
| CN114000907A (en) * | 2021-12-10 | 2022-02-01 | 重庆邮电大学 | Mine ventilation equipment intelligent regulation and control system based on digital twin technology |
| CN114564259A (en) * | 2022-01-24 | 2022-05-31 | 杭州博联智能科技股份有限公司 | Method and system for generating visual interface |
| CN114637262B (en) * | 2022-03-10 | 2022-11-15 | 天津科技大学 | Decision control method and system of intelligent factory digital twin information based on 5G drive |
| CN114738031B (en) * | 2022-04-15 | 2023-04-14 | 北京科技大学 | Natural and mechanical ventilation coupled construction tunnel digital twin intelligent ventilation method |
| CN114925586B (en) * | 2022-04-15 | 2023-06-16 | 中煤科工集团重庆研究院有限公司 | Parameter calculation method for long-pressure short-pumping wind-control dust-removing equipment of fully-mechanized coal mining face |
| 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 |
| CN115587542B (en) * | 2022-11-02 | 2023-06-16 | 四川奥凸水处理系统工程有限公司 | Groundwater inversion simulation method, system, equipment and medium based on reinforcement learning |
| CN115985491B (en) * | 2023-03-21 | 2023-07-18 | 安徽通灵仿生科技有限公司 | Medical information processing system of interventional ventricular catheter pump |
| CN117404735B (en) * | 2023-12-07 | 2024-04-02 | 湖北华中电力科技开发有限责任公司 | Intelligent air quality early warning system for limited space safety operation |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH11107700A (en) * | 1997-09-30 | 1999-04-20 | Mitsubishi Electric Corp | Control system using virtual-actual picture simulation |
| CN104018884B (en) * | 2014-06-13 | 2017-02-01 | 武汉理工大学 | Intelligent sensing and remote and unified regulating system for deep-mining mine ventilation |
| JP7254819B2 (en) * | 2018-02-05 | 2023-04-10 | ジール・アベッグ エスエー | How to optimize fan efficiency and/or operating performance or fan placement |
| CN110905580B (en) * | 2019-11-30 | 2021-07-27 | 西安科技大学 | Fully-mechanized excavation face ventilation optimization method |
| CN111177942B (en) * | 2020-01-06 | 2023-04-18 | 中国矿业大学(北京) | Digital twin intelligent monitoring system for unmanned fully-mechanized excavation working face of mine |
| CN112332444A (en) * | 2020-09-14 | 2021-02-05 | 华北电力大学（保定） | Microgrid energy management system based on digital twins |
| CN112392485B (en) * | 2020-11-12 | 2021-08-17 | 临沂矿业集团菏泽煤电有限公司 | Transparent digital twin self-adaptive mining system and method for fully mechanized coal mining face |
-
2021
- 2021-07-08 CN CN202110773534.8A patent/CN113356916B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN113356916A (en) | 2021-09-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113356916B (en) | Mine air flow regulation and control virtual system based on digital twin technology and intelligent regulation and control method | |
| Peng et al. | A self-learning dynamic path planning method for evacuation in large public buildings based on neural networks | |
| CN107463751B (en) | Crowd grouping evacuation simulation method and system based on binary DBSCAN clustering algorithm | |
| KR20200144998A (en) | Computing system for virtual sensor implementation using digital twin and method for realtime data collection thereof | |
| Zheng et al. | A generative architectural and urban design method through artificial neural networks | |
| CN109684714B (en) | Building design method based on machine learning and BIM technology | |
| WO2019076044A1 (en) | Mobile robot local motion planning method and apparatus and computer storage medium | |
| CN106951990A (en) | Electric load intelligent Forecasting and device | |
| CN110930514B (en) | Three-dimensional grid model self-adaptive simplification method based on local region characteristics | |
| Xie et al. | A virtual test and evaluation method for fully mechanized mining production system with different smart levels | |
| Hu et al. | Predicting crowd egress and environment relationships to support building design optimization | |
| Lei et al. | A comparative study of six hybrid prediction models for uniaxial compressive strength of rock based on swarm intelligence optimization algorithms | |
| Liu et al. | Automated clash resolution for reinforcement steel design in precast concrete wall panels via generative adversarial network and reinforcement learning | |
| Zeng et al. | Artificial Intelligence tool for fire safety design (IFETool): Demonstration in large open spaces | |
| CN116542185A (en) | Digital twin pipe gallery system based on reduced order simulation model and real-time calibration algorithm | |
| CN103529803A (en) | Advanced control experiment system for cloth sintering and segregation on basis of cloud virtualization | |
| Xiong et al. | ShapeArchit: shape-inspired architecture design with space planning | |
| CN108595775B (en) | The optimization method and system of barrier geometry during crowd evacuation emulation | |
| CN115509178A (en) | Digital twin driven cutter wear monitoring method and numerical control machine tool equipment | |
| CN116029331A (en) | Big data crowd sensing collaborative prediction method for safety state of coal mine working face | |
| CN115964503A (en) | Safety risk prediction method and system based on community equipment facilities | |
| Li et al. | A study on urban block design strategies for improving pedestrian-level wind conditions: CFD-based optimization and generative adversarial networks | |
| Wang et al. | Data-driven simulation of pedestrian movement with artificial neural network | |
| Xu | Risk Assessment of Green Intelligent Building Based on Artificial Intelligence | |
| He et al. | A data-driven method for dynamic load forecasting of scraper conveyer based on rough set and multilayered self-normalizing gated recurrent network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |