Coal mining work information fusion and intelligent decision system under high gas condition
By setting up a sensor network and information fusion model in underground coal mines under high gas conditions, the operating parameters of the coal mining machine and ventilation fan can be adjusted in real time, solving the problem of poor gas concentration control and achieving safe and efficient coal mining operations.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SHANXI LUAN MINING GRP
- Filing Date
- 2024-01-22
- Publication Date
- 2026-07-07
AI Technical Summary
Existing technologies make it difficult to achieve real-time coordination between coal mining operations and ventilation equipment in underground coal mines under high gas conditions, resulting in poor gas concentration control and potential safety hazards.
The system employs mine gas acquisition sensors, coal mining machine data acquisition sensors, ventilation fan data acquisition sensors, data acquisition devices, PC terminals, cloud terminals, and PLCs. Data transmission and fusion are achieved through a 5G network. The system utilizes an information fusion model to adjust the operating parameters of the coal mining machine and ventilation fan in real time, enabling intelligent decision-making and closed-loop control.
It enables real-time control of gas concentration in coal mining under high gas conditions, ensuring a balance between coal mining speed and gas emission, and improving the real-time performance and safety of the ventilation fan.
Smart Images
Figure CN117759242B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of intelligent coal mining technology, and in particular to an information fusion and intelligent decision-making system for coal mining under high gas conditions. Background Technology
[0002] Information fusion technology is an information processing process that uses computer technology to automatically analyze and synthesize observation information from several sensors acquired in a time sequence under certain criteria, in order to complete the required decision-making and estimation tasks. Intelligent decision-making utilizes human knowledge and, with the help of computers, uses artificial intelligence methods to solve complex decision problems.
[0003] High-gas mines refer to mines where, during the mining process, coal seams release high concentrations of gases containing hydrogen, methane, ethylene, carbon monoxide, nitrogen dioxide, and naphtha. Ventilation systems are not only fundamental to the smooth operation of coal mining but also prevent gas accidents during high-gas mining operations.
[0004] Currently, due to the complex underground environment of coal mines, the deployment of sensors and the acquisition of operational data for underground equipment are still in their infancy. Therefore, information fusion technology and intelligent decision-making technology are rarely applied in underground coal mines. Coal mine ventilation equipment mainly consists of ventilation fans. The operating power and airflow of these fans are typically monitored, and their operating status is adjusted according to the mine's gas emission rate. However, the gas emission rate is not constant and is closely related to the mining speed of the coal face mining machine and the coal seam distribution. In high-gas mining operations, the key lies in the coordination between the ventilation equipment and the mining equipment to ensure that the underground gas concentration remains within a reasonable range. Existing technologies cannot effectively address the impact between mining operations and gas emission rates, and the real-time performance of adjusting the ventilation fan's operating status needs improvement. Summary of the Invention
[0005] To address the aforementioned technical problems, this invention provides an information fusion and intelligent decision-making system for coal mining under high-gas conditions. The technical solution of this invention is as follows:
[0006] A coal mining information fusion and intelligent decision-making system under high gas conditions includes a mine gas acquisition sensor, a coal mining machine data acquisition sensor, a ventilation fan data acquisition sensor, a data acquisition unit, a PC terminal, a cloud terminal, and a PLC. The mine gas acquisition sensor, the coal mining machine data acquisition sensor, and the ventilation fan data acquisition sensor are wirelessly connected to the data acquisition unit. The data acquisition unit is connected to both the PC terminal and the cloud terminal via a 5G network. The PC terminal is connected to the PLC, and the PLC is connected to both the coal mining machine and the ventilation fan.
[0007] The mine gas acquisition sensor, coal mining machine data acquisition sensor, and ventilation fan data acquisition sensor respectively collect the current mine gas concentration, current coal mining machine operation data, and current ventilation fan operation data, and transmit them wirelessly to the data acquisition unit. The data acquisition unit transmits the current mine gas concentration, current coal mining machine operation data, and current ventilation fan operation data to the cloud and PC via a 5G network. The cloud classifies and saves the received data as historical operating condition data for later use in training the information fusion model. The PC fuses the current mine gas concentration, current coal mining machine operation data, and current ventilation fan operation data, and makes intelligent decisions on the operating parameters of the coal mining machine and ventilation fan based on the fusion results. The intelligent decision results are then transmitted to the PLC, which adjusts the operating status of the coal mining machine and ventilation fan in real time based on the intelligent decision results.
[0008] Optionally, when the PC performs data fusion on the current mine gas concentration, current coal mining machine operation data, and current ventilation fan operation data, and makes intelligent decisions on the operating parameters of the coal mining machine and ventilation fan based on the fusion results, it includes:
[0009] On the PC, a pre-trained information fusion model is used to fuse the current mine gas concentration, the current coal mining machine operation data, and the current ventilation fan operation data to obtain the current coal mining status risk value. Based on the current coal mining status risk value, intelligent decisions are made on the operating parameters of the coal mining machine and the ventilation fan.
[0010] Optionally, the information fusion model includes a channel attention network, a convolutional neural network, and a long short-term memory network connected in sequence; when the PC performs data fusion on the current mine gas concentration, current coal mining machine operation data, and current ventilation fan operation data through the pre-trained information fusion model to obtain the current coal mining status risk value, it includes:
[0011] The channel attention network assigns weights to the current mine gas concentration, the current coal mining machine operation data, and the current ventilation fan operation data, resulting in weighted mine gas concentration, coal mining machine operation data, and ventilation fan operation data.
[0012] The convolutional neural network extracts low-level features from the assigned mine gas concentration, the assigned coal mining machine operation data, and the assigned ventilation fan operation data to obtain low-level features;
[0013] Low-level features are input into a Long Short-Term Memory (LSTM) network for high-level feature extraction, and the LSM network outputs the current coal mining status risk value.
[0014] Optionally, when the PC makes intelligent decisions on the operating parameters of the coal mining machine and the ventilation fan based on the fusion results, it includes:
[0015] When the risk value of the coal mining status is less than 0.2, the PC terminal determines that the operating parameters of the coal mining machine and the ventilation fan will not be adjusted.
[0016] When the risk value of the coal mining status is greater than 0.2 and less than 0.5, the PC terminal determines to increase the speed of the ventilation fan by 10%, decrease the speed of the left and right drums of the coal mining machine by 20%, and decrease the travel speed of the traction unit of the coal mining machine by 10%.
[0017] When the risk value of the coal mining status is greater than 0.5 and less than 0.8, the PC terminal determines to increase the speed of the ventilation fan by 30%, decrease the speed of the left and right drums of the coal mining machine by 40%, and decrease the travel speed of the traction unit of the coal mining machine by 25%.
[0018] When the risk value of the coal mining status is greater than 0.8, the PC terminal determines to increase the speed of the ventilation fan by 50% and stop the coal mining machine for inspection.
[0019] Optionally, after the PC determines to reduce the speed of the left and right drums of the coal mining machine and reduce the travel speed of the traction unit of the coal mining machine, the following steps are also included:
[0020] Step 1: The PC terminal obtains the preset action event set for the coal mining machine and ventilation fan, and determines the time relationship constraint matrix corresponding to the action event set based on the preset basic time relationship of the events.
[0021] Step 2: The PC simplifies the time relationship constraint matrix to merge simultaneous time points on the time axis; calculates the basic assignments of the simplified time relationship constraint matrix, and the set of all basic assignments forms an assignment directed graph, which is the set of all feasible time relationships for each action event in the action event set;
[0022] Step 3: On the PC, assign a duration to each action event in the action event set, and calculate the total running time of all action events in the action event set corresponding to each basic assignment in the assigned directed graph. Select the shortest total running time as the execution time of the action event set.
[0023] Optionally, when the PC simplifies the time relationship constraint matrix, it includes:
[0024] If there is no element Q(x,y) = {0} in the time constraint matrix Q, and Q(x,y) represents the element in the x-th row and y-th column of the time constraint matrix Q, then the time constraint matrix Q does not need to be simplified.
[0025] If there exists an element Q(x,y) = {0} in the time constraint matrix Q, let J(Q) denote the set of row numbers of the rows in Q, and take x,y∈J(Q) such that Q(x,y) = {0}; ,make , If the above conditions are met, then the x-th and y-th rows of the time constraint matrix Q are deleted and the x′-th row is added to achieve row simplification of the time constraint matrix Q; the same method is used to achieve column simplification of the time constraint matrix; after row simplification and column simplification of the time constraint matrix Q, the simplified time constraint matrix is obtained.
[0026] Optionally, the mine gas acquisition sensor includes a gas sensor for detecting gas concentration data at the mine working face, and the gas sensor is installed on the main body of the coal mining machine;
[0027] The data acquisition sensors for the coal mining machine include an encoder for detecting the rotational speed of the coal mining machine drum, an inclination sensor for detecting the tilt angle of the coal mining machine rocker arm, and an odometer for detecting the speed of the traction unit of the coal mining machine. The encoder is installed on the drum shaft of the coal mining machine, the inclination sensor is installed on the rocker arm of the coal mining machine, and the odometer is installed on the wheel of the traction unit of the coal mining machine.
[0028] The data acquisition sensor for the ventilator includes a flow meter for detecting the air volume of the ventilator and a power meter for detecting the power of the ventilator. The flow meter is installed on the ventilation duct of the ventilator, and the power meter is installed on the motor of the ventilator.
[0029] All of the above-mentioned optional technical solutions can be combined arbitrarily, and the present invention will not provide a detailed description of the structure after each combination.
[0030] By means of the above solution, the beneficial effects of the present invention are as follows:
[0031] By setting up mine gas acquisition sensors, coal mining machine data acquisition sensors, ventilation fan data acquisition sensors, data acquisition devices, PC terminals, cloud computing, and PLCs, intelligent decision-making and adjustments to the operating parameters of the coal mining machine and ventilation fans can be made based on the detected mine gas concentration, coal mining machine operation data, and ventilation fan operation data. This links the mining speed with gas emission and ventilation fan airflow under high-gas conditions, ensuring that the underground gas concentration in the coal mine remains within a reasonable range while enabling real-time adjustment of the ventilation fan's operating status. Furthermore, by using real-time data detected by sensors on the coal mining machine and ventilation fans to adjust their operating status in real time, closed-loop control of coal mining operations under high-gas conditions is achieved.
[0032] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, the preferred embodiments of the present invention are described in detail below with reference to the accompanying drawings. Attached Figure Description
[0033] Figure 1 This is a schematic diagram of the composition structure of the present invention.
[0034] Figure 2 This is a schematic diagram of the components of the information fusion model.
[0035] Figure 3 It is a pre-defined diagram of the basic time relationships of events. Detailed Implementation
[0036] The specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and are not intended to limit the scope of the invention.
[0037] like Figure 1 As shown, the coal mining information fusion and intelligent decision-making system under high gas conditions provided by the present invention includes a mine gas acquisition sensor, a coal mining machine data acquisition sensor, a ventilation fan data acquisition sensor, a data acquisition unit, a PC terminal, a cloud terminal, and a PLC. The mine gas acquisition sensor, the coal mining machine data acquisition sensor, and the ventilation fan data acquisition sensor are connected to the data acquisition unit wirelessly. The data acquisition unit is connected to both the PC terminal and the cloud terminal via a 5G network. The PC terminal is connected to the PLC via a cable, and the PLC is connected to both the coal mining machine and the ventilation fan via cables.
[0038] The system consists of a mine gas acquisition sensor, a coal mining machine data acquisition sensor, and a ventilation fan data acquisition sensor. These sensors collect the current mine gas concentration, the current coal mining machine operating data, and the current ventilation fan operating data, respectively. The data is then transmitted wirelessly to a data acquisition unit. The data acquisition unit transmits these data to the cloud and a PC via a 5G network. The cloud classifies and saves the received data as historical data for later use in training the information fusion model. The PC fuses the current mine gas concentration, coal mining machine operating data, and ventilation fan operating data. Based on the fusion results, the PC makes intelligent decisions on the operating parameters of the coal mining machine and ventilation fan. The intelligent decision results are then transmitted to a PLC. The PLC adjusts the operating status of the coal mining machine and ventilation fan in real time based on the intelligent decision results.
[0039] In this embodiment of the invention, the mine gas acquisition sensor, the coal mining machine data acquisition sensor, and the ventilation fan data acquisition sensor are connected to the data acquisition unit wirelessly. This avoids the problems associated with traditional wired transmission methods, such as line damage, sparking, and significant property losses and personnel safety threats under high gas conditions, which can occur due to the complex underground environment. This approach is more suitable for underground coal mines and has greater applicability. The cloud-based system categorizes and stores the received data as historical operating data for later use, forming a complete equipment history database. This database facilitates later verification and innovative research, and can also be used to train information fusion models.
[0040] Optionally, the mine gas acquisition sensor includes a gas sensor for detecting gas concentration data at the mine working face; the gas sensor is installed on the main body of the coal mining machine; the coal mining machine data acquisition sensor includes an encoder for detecting the rotational speed of the coal mining machine drum, an inclination sensor for detecting the tilt angle of the coal mining machine rocker arm, and an odometer for detecting the speed of the coal mining machine traction unit; the encoder is installed on the drum shaft of the coal mining machine, the inclination sensor is installed on the rocker arm of the coal mining machine, and the odometer is installed on the wheel of the traction unit of the coal mining machine; the ventilation fan data acquisition sensor includes a flow meter for detecting the air volume of the ventilation fan and a power meter for detecting the power of the ventilation fan; the flow meter is installed on the ventilation pipe of the ventilation fan, and the power meter is installed on the motor of the ventilation fan.
[0041] To ensure smooth data fusion, each sensor can be configured to collect data synchronously. When the PC-side fuses the current mine gas concentration, current coal mining machine operating data, and current ventilation fan operating data, and makes intelligent decisions regarding the operating parameters of the coal mining machine and ventilation fan based on the fusion results, it can use a pre-trained information fusion model to fuse the current mine gas concentration, current coal mining machine operating data, and current ventilation fan operating data to obtain the current coal mining status risk value. Based on this risk value, intelligent decisions are then made regarding the operating parameters of the coal mining machine and ventilation fan.
[0042] The embodiments of this invention do not specifically limit the composition and structure of the information fusion model. In specific implementations, such as... Figure 2 As shown, the information fusion model may include a channel attention network, a convolutional neural network, and a long short-term memory network connected in sequence. Based on this, when the PC performs data fusion on the current mine gas concentration, current coal mining machine operation data, and current ventilation fan operation data using the pre-trained information fusion model to obtain the current coal mining status risk value, it can be achieved as follows: The channel attention network assigns weights to the current mine gas concentration, current coal mining machine operation data, and current ventilation fan operation data, obtaining weighted mine gas concentration, weighted coal mining machine operation data, and weighted ventilation fan operation data; the convolutional neural network performs low-level feature extraction on the weighted mine gas concentration, weighted coal mining machine operation data, and weighted ventilation fan operation data, obtaining low-level features; the low-level features are input into the long short-term memory network for high-level feature extraction, and the long short-term memory network outputs the current coal mining status risk value.
[0043] In specific implementation, taking the entire system including the above six sensors as an example, each is used as one of the six channels for the information fusion model input, and the size of the input data of the information fusion model is set to 6*1024 (number of data). The channel attention network assigns weights to the data input from the six channels. The convolutional neural network sets two convolutional pooling operations with the following parameters: (1) 6 input channels, 16 output channels, 30 convolution kernel size, 2 convolution stride, ReLU activation function, 2 pooling kernel size for max pooling, and 2 pooling stride; (2) 16 input channels, 6 output channels, 15 convolution kernel size, 2 convolution stride, ReLU activation function, 2 pooling kernel size for max pooling, and 2 pooling stride. The final low-level feature size obtained by the convolutional neural network is 6*59. The input size of the long short-term memory network is 6, the hidden layer size is 64, and the output size is 1, representing the current coal mining risk value. The closer the current coal mining risk value is to 1, the higher the risk.
[0044] Specifically, when the PC makes intelligent decisions on the operating parameters of the coal mining machine and the ventilation fan based on the fusion results, it can achieve the following strategies: When the coal mining status risk value is less than 0.2, the PC determines not to adjust the operating parameters of the coal mining machine and the ventilation fan; when the coal mining status risk value is greater than 0.2 and less than 0.5, the PC determines to increase the ventilation fan speed by 10%, decrease the speed of the left and right drums of the coal mining machine by 20%, and decrease the travel speed of the traction unit of the coal mining machine by 10%; when the coal mining status risk value is greater than 0.5 and less than 0.8, the PC determines to increase the ventilation fan speed by 30%, decrease the speed of the left and right drums of the coal mining machine by 40%, and decrease the travel speed of the traction unit of the coal mining machine by 25%; when the coal mining status risk value is greater than 0.8, the PC determines to increase the ventilation fan speed by 50% and stop the coal mining machine for inspection.
[0045] Furthermore, the operating parameters of the coal mining machine and the ventilation fan correspond to the operating time. That is, in the intelligent decision-making process for the operation of the coal mining machine and the ventilation fan, both the operating parameters and the operating time must be determined. When the above-mentioned operating parameter adjustment strategy reduces the rotational speed of the left and right drums of the coal mining machine and the traveling speed of the traction unit of the coal mining machine, the operating time of each action event of the coal mining machine will be appropriately extended. In order to reasonably adjust the operating time, the embodiments of the present invention also include the following content on intelligent decision-making regarding the operating time of action events.
[0046] Specifically, after the PC terminal determines to reduce the rotational speed of the left and right drums of the coal mining machine and to reduce the traveling speed of the traction unit of the coal mining machine, the embodiments of the present invention further include the following steps:
[0047] Step 1: The PC terminal obtains the preset action event set for the coal mining machine and ventilation fan, and determines the time relationship constraint matrix corresponding to the action event set based on the preset basic time relationship of the events.
[0048] When controlling the operation of the coal mining machine and the ventilation fan, multiple sets of action events are preset and executed in sequence. This embodiment of the invention defines a single operation process of the coal mining machine as including six action events, namely: the coal mining machine left drum (start / stop), the coal mining machine right drum (start / stop), the coal mining machine left rocker arm (start / stop), the coal mining machine right rocker arm (start / stop), the coal mining machine traction unit travel (start / stop), and the ventilation fan ventilation (start / stop).
[0049] The preset basic time relationship of events is as follows: Figure 3 As shown, any two action events contain, as Figure 3 The 13 temporal relationships shown above, and the 6 action events mentioned above, can be used... Figure 3 This refers to one or more of the 13 defined time relationships. For example, suppose the left drum (start / stop) of the coal mining machine is action event I, and the right drum (start / stop) is action event J. According to the coal mining machine operation procedures, the time relationship between action events I and J can be: <, m, o, fi, di, s, =, si, d, f, oi, mi, and >, with corresponding time point relationships as follows: , , , , , , , , , , , , .
[0050] The execution time interval of each action event is represented by the start time and end time. Indicates the start time of action event I. This indicates the end time of action event I. Similarly, This indicates the start time of event J. This represents the end time of action event J. Based on this, the basic temporal relationship between any two action events can be represented as follows: the starting time point partition set A = {-1, 0, 1}, where -1, 0, and 1 represent the intervals (-∞, a), [a], and (a, ∞) respectively; similarly, the ending time point B = {-1, 0, 1}, where -1, 0, and 1 represent the intervals (-∞, b), [b], and (b, ∞) respectively. The temporal relationship constraint between two action events I and J is represented by the product of sets, i.e., R(I, J) = (A1(i, j) × A2(i, j)) × (B1(i, j) × B2(i, j)), where action event I = [a... i bi ], Action event J = [a j b j A1(i,j) is a i With a j The constraints between them, i.e., a j Falling to a i The interval; A2(i,j) is a i With b j The constraint between them, i.e., b j Falling to a i The interval; B1(i,j) is the interval of b. i With a j The constraints between them, i.e., a j Landing on b i The interval; B2(i,j) is the interval of b. i With b j The constraint between them, i.e., b j Landing on b i The time relationship constraint between any two action events I and J can be represented by a 2x2 matrix, i.e.:
[0051] .
[0052] Based on the above, the time relationship constraint matrix Q corresponding to the action event set is determined according to the preset basic time relationship of events as follows:
[0053] ;
[0054] Where, = represents , , ,~ indicates that two action events satisfy any one of the basic event relations.
[0055] Step two: The PC simplifies the time relationship constraint matrix to merge simultaneously occurring time points on the time axis; it calculates the basic assignments of the simplified time relationship constraint matrix, and the set of all basic assignments forms an assignment directed graph, which is the set of all feasible time relationships for each action event in the action event set. A basic assignment is the combination of the start and end times of each action event in the action event set.
[0056] Specifically, when the PC simplifies the time relationship constraint matrix, there are two cases:
[0057] Case 1: If there is no element Q(x,y) = {0} in the time constraint matrix Q, where Q(x,y) represents the element in the x-th row and y-th column of the time constraint matrix Q, then the time constraint matrix Q does not need to be simplified.
[0058] The second case: If there exists an element Q(x, y) = {0} in the time relationship constraint matrix Q, let J(Q) denote the set of row numbers of the rows of Q. Take x, y ∈ J(Q) and Q(x, y) = {0}; , let , , if the above conditions are satisfied, then delete the x-th and y-th rows in the time relationship constraint matrix Q and add the x'-th row to achieve row reduction of the time relationship constraint matrix Q; use the same method to achieve column reduction of the time relationship constraint matrix Q; after the time relationship constraint matrix Q undergoes row reduction and column reduction respectively, the reduced time relationship constraint matrix is obtained.
[0059] Simplifying the time relationship constraint matrix can merge the time points that occur simultaneously on the time axis into one point, thereby reducing the dimension of the time relationship constraint matrix and reducing the complexity of subsequent calculations.
[0060] Specifically, when the PC calculates the basic assignment of the reduced time relationship constraint matrix and constructs an assignment digraph using the basic assignment, it can be achieved through assignment digraph connection operations, time relationship constraint right shift operations, and time width constraint right shift operations.
[0061] In specific implementation, when calculating the basic assignment F(d) of the reduced time relationship constraint matrix, let F(d) = (d(t1), d(t2), …, d(t 2n )) be an assignment, where T = (t1, t2, …, t 2n ) and satisfies the following conditions: (1) , if d(x) < d(y), then -1 ∈ Q(x, y); (2) , let Δi = a i - b i , then there is Δi min < Δi < Δi max , then F(d) is a basic assignment of the reduced time relationship constraint matrix, and combining each basic assignment F(d) forms an assignment digraph. Δi min and Δi max are respectively the minimum and maximum values of the execution time width in the time width data samples of the action event I. The elements in F(d) are grouped in pairs, respectively representing the start time point and end time point of a certain action event.
[0062] For example, if in the set of action events, it is specified that among 6 action events, the ventilator starts first and closes last, then in a basic assignment F(d) = (d(t1), d(t2), …, d(t 12 )), d(t 11 ) is 1 and d(t 12 ) is 12.
[0063] Step 3: On the PC, assign a duration to each action event in the action event set, and calculate the total running time of all action events in the action event set corresponding to each basic assignment in the assigned directed graph. Select the shortest total running time as the execution time of the action event set.
[0064] Specifically, when assigning duration to each action event in the action event set, the duration of each action event is determined based on experience, and this determined empirical duration is used as the assigned duration of each action event in the action event set. Furthermore, after determining the total running time of all action events corresponding to each basic assignment in the directed graph, the shortest total running time is selected as the execution time of the action event set. Subsequently, when determining the running time of each action event in the action event set controlling the coal mining machine and ventilation fan, each action event in the action event set is executed according to the time node corresponding to the selected basic assignment.
[0065] When the PLC adjusts the operating status of the coal mining machine and ventilation fan in real time based on the intelligent decision-making results, it does so by sending new operating parameter commands to the coal mining machine and ventilation fan. Upon receiving the new operating parameters, the coal mining machine and ventilation fan change the drum speed, rocker arm angle, traction unit speed, and ventilation fan speed in real time, thereby linking the mining speed under high-gas conditions with the gas emission rate and ventilation fan airflow. Furthermore, after the coal mining machine and ventilation fan adjust their operating parameters, the mine gas acquisition sensors, coal mining machine data acquisition sensors, and ventilation fan data acquisition sensors continue to collect data in real time and continue the above process, thus achieving closed-loop control of coal mining operations under high-gas conditions. In summary, this invention can intelligently determine the operating parameters of the coal mining machine and ventilation fan based on the detected working face gas concentration, and achieve closed-loop control of coal mining operations under high-gas conditions through data detected by sensors on the coal mining machine and ventilation fan equipment.
[0066] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.
Claims
1. A coal mining information fusion and intelligent decision-making system under high gas conditions, characterized in that, The system includes a mine gas acquisition sensor, a coal mining machine data acquisition sensor, a ventilation fan data acquisition sensor, a data acquisition unit, a PC terminal, a cloud terminal, and a PLC. The mine gas acquisition sensor, the coal mining machine data acquisition sensor, and the ventilation fan data acquisition sensor are connected to the data acquisition unit wirelessly. The data acquisition unit is connected to both the PC terminal and the cloud terminal via a 5G network. The PC terminal is connected to the PLC, and the PLC is connected to both the coal mining machine and the ventilation fan. The mine gas acquisition sensor, coal mining machine data acquisition sensor, and ventilation fan data acquisition sensor respectively collect the current mine gas concentration, current coal mining machine operation data, and current ventilation fan operation data, and then transmit them wirelessly to the data acquisition unit. The data acquisition unit transmits the current mine gas concentration, current coal mining machine operation data, and current ventilation fan operation data to the cloud and PC via a 5G network. The cloud classifies and saves the received data as historical working condition data for later use in training the information fusion model. The PC fuses the current mine gas concentration, current coal mining machine operation data, and current ventilation fan operation data, and makes intelligent decisions on the operating parameters of the coal mining machine and ventilation fan based on the fusion results. The intelligent decision results are then transmitted to the PLC, and the PLC adjusts the operating status of the coal mining machine and ventilation fan in real time based on the intelligent decision results. The PC-side system uses a pre-trained information fusion model to fuse current mine gas concentration, current coal mining machine operation data, and current ventilation fan operation data to obtain the current coal mining status risk value. Based on this risk value, it makes intelligent decisions regarding the operating parameters of the coal mining machine and ventilation fan. The information fusion model comprises a channel attention network, a convolutional neural network, and a long short-term memory network connected in sequence. The channel attention network assigns weights to the current mine gas concentration, coal mining machine operation data, and ventilation fan operation data, resulting in weighted mine gas concentration, coal mining machine operation data, and ventilation fan operation data. The convolutional neural network extracts low-level features from these data, obtaining low-level features. These low-level features are then input into the long short-term memory network for high-level feature extraction, and the long short-term memory network outputs the current coal mining status risk value. Specifically: when the coal mining status risk value is less than 0.2, the PC terminal determines not to adjust the operating parameters of the coal mining machine and the ventilation fan; when the coal mining status risk value is greater than 0.2 and less than 0.5, the PC terminal determines to increase the ventilation fan speed by 10%, decrease the left and right drum speeds of the coal mining machine by 20%, and decrease the traction speed of the coal mining machine by 10%; when the coal mining status risk value is greater than 0.5 and less than 0.8, the PC terminal determines to increase the ventilation fan speed by 30%, decrease the left and right drum speeds of the coal mining machine by 40%, and decrease the traction speed of the coal mining machine by 25%; when the coal mining status risk value is greater than 0.8, the PC terminal determines to increase the ventilation fan speed by 50% and stop the coal mining machine for inspection.
2. The information fusion and intelligent decision-making system for coal mining under high gas conditions according to claim 1, characterized in that, After the PC terminal determines to reduce the speed of the left and right drums of the coal mining machine and reduce the travel speed of the traction unit of the coal mining machine, it also includes: Step 1: The PC terminal obtains the preset action event set for the coal mining machine and ventilation fan, and determines the time relationship constraint matrix corresponding to the action event set based on the preset basic time relationship of the events. Step 2: The PC simplifies the time relationship constraint matrix to merge simultaneous time points on the time axis; calculates the basic assignments of the simplified time relationship constraint matrix, and the set of all basic assignments forms an assignment directed graph, which is the set of all feasible time relationships for each action event in the action event set; Step 3: On the PC, assign a duration to each action event in the action event set, and calculate the total running time of all action events in the action event set corresponding to each basic assignment in the assigned directed graph. Select the shortest total running time as the execution time of the action event set.
3. The information fusion and intelligent decision-making system for coal mining under high gas conditions according to claim 2, characterized in that, When the PC simplifies the time relationship constraint matrix, it includes: If there is no element Q(x, y) = {0} in the time constraint matrix Q, and Q(x, y) represents the element in the x-th row and y-th column of the time constraint matrix Q, then the time constraint matrix Q does not need to be simplified. If there exists an element Q(x,y) = {0} in the time constraint matrix Q, let J(Q) denote the set of row numbers of the rows of Q, and take x,y∈J(Q) and Q(x,y) = {0}; ,make , If the above conditions are met, then the x-th and y-th rows of the time constraint matrix Q are deleted and the x′-th row is added to achieve row simplification of the time constraint matrix Q; the same method is used to achieve column simplification of the time constraint matrix; after row simplification and column simplification of the time constraint matrix Q, the simplified time constraint matrix is obtained.
4. The information fusion and intelligent decision-making system for coal mining under high gas conditions according to claim 1, characterized in that, The mine gas acquisition sensor includes a gas sensor for detecting gas concentration data at the mine working face, and the gas sensor is installed on the main body of the coal mining machine; The data acquisition sensors for the coal mining machine include an encoder for detecting the rotational speed of the coal mining machine drum, an inclination sensor for detecting the tilt angle of the coal mining machine rocker arm, and an odometer for detecting the speed of the traction unit of the coal mining machine. The encoder is installed on the drum shaft of the coal mining machine, the inclination sensor is installed on the rocker arm of the coal mining machine, and the odometer is installed on the wheel of the traction unit of the coal mining machine. The data acquisition sensor for the ventilator includes a flow meter for detecting the air volume of the ventilator and a power meter for detecting the power of the ventilator. The flow meter is installed on the ventilation duct of the ventilator, and the power meter is installed on the motor of the ventilator.