Coal face roof disaster monitoring method and system based on edge calculation

Abstract

The invention relates to the technical field of coal mining, and provides a coal face roof disaster monitoring method and system based on edge calculation. According to the method, a plurality of pressure radio frequency sensors are uniformly arranged on a hydraulic support of a coal face so as to monitor and wirelessly transmit top plate pressure data of the coal face; meanwhile, an area edge gateway is arranged on an air inlet roadway of the coal face, and roof pressure data monitored by the pressure radio frequency sensor are received; then, marking the roof pressure data within the time period [t1, t2] received by the area edge gateway, and determining an optimal hyperplane according to the marked roof pressure marking data; and finally, determining the vector direction from the roof pressure data received by the area edge gateway within the time period [t3, t4] to the optimal hyperplane, classifying the received roof pressure data within the time period [t3, t4], and determining the roof disaster and danger area of the coal face in the production process according to the classification result.

Description

Technical Field

[0001] The present application relates to the field of coal mining technology, and in particular to a method and system for monitoring roof disasters in a coal mining face based on edge computing. Background Art

[0002] Coal mine roadway roof disasters refer to the situation in which the roadway roof breaks, collapses or deforms due to geological conditions, engineering structures and other factors during underground coal mining. These disasters will pose a serious threat to mine production safety and the safety of workers. In order to prevent and reduce the occurrence of coal mine roadway roof disasters, effective geological surveys, roadway support, gas extraction, safety monitoring and other measures are needed to improve the safety level of mine production.

[0003] At present, when monitoring roof disasters, the collection of indicator data usually adopts wired transmission. The data needs to go through cumbersome processes such as monitoring substations and underground ring networks before it can be uploaded to the ground for analysis, which seriously affects the real-time and timely response capabilities of the data. Moreover, the existing monitoring methods lack sufficient theoretical support, and monitoring is carried out throughout the entire cycle of coal mining working face recovery, and data cannot be collected scientifically, accurately and effectively.

[0004] Therefore, there is an urgent need to provide a technical solution to the above-mentioned deficiencies in the prior art. Summary of the invention

[0005] The purpose of this application is to provide a method and system for monitoring roof disasters in a coal mining face based on edge computing to solve or alleviate the problems existing in the above-mentioned prior art.

[0006] In order to achieve the above objectives, this application provides the following technical solutions:

[0007] The present application provides a method for monitoring roof disasters of a coal mining face based on edge computing, comprising: step S101, evenly arranging a plurality of pressure radio frequency sensors on the hydraulic support of the coal mining face to monitor the roof pressure data of the coal mining face and wirelessly transmitting the data through a built-in LoRa module; and arranging a regional edge gateway in the air intake lane of the coal mining face to receive the roof pressure data monitored by the pressure radio frequency sensor; step S102, the regional edge gateway marks the received roof pressure data within the time period [t1, t2], and determines an optimal hyperplane according to the obtained roof pressure mark data; step S103, determining a vector direction from the roof pressure data within the time period [t3, t4] received by the regional edge gateway to the optimal hyperplane, so as to classify the received roof pressure data within the time period [t3, t4], and determine the roof disaster risk area of ​​the coal mining face during the production process according to the classification result; wherein the time period [t3, t4] is later than the time period [t1, t2].

[0008] Preferably, in step S102, the received roof pressure data in the time period [t1, t2] is marked based on a clustering algorithm to obtain the roof pressure marked data, and the optimal hyperplane is determined according to the roof pressure marked data.

[0009] Preferably, the step of marking the received top plate pressure data within the time period [t1, t2] based on a clustering algorithm comprises:

[0010] According to the formula:

[0011]

[0012] Determine the top plate pressure data x received in the time period [t1, t2] of the i-th i To the predetermined mth cluster center c m Wherein, n is the number of the top plate pressure data in the time period [t1, t2]; n is a positive integer; m=2;

[0013] According to the top plate pressure data x i To the mth cluster center c m The distance d(i, m) of the top plate pressure data x i Perform clustering level labeling.

[0014] Preferably, determining the optimal hyperplane based on the roof pressure marking data includes: expanding the roof pressure marking data in a three-dimensional space coordinate system with working face advancement distance, hydraulic support number and pressure value as characteristic indicators, and determining the optimal hyperplane based on a preset objective function.

[0015] Preferably, determining the optimal hyperplane based on a preset objective function includes: based on the objective function:

[0016] max margin(W, p)

[0017] Determine the optimal hyperplane; wherein W is the direction vector of the hyperplane obtained after the n top plate pressure mark data are expanded in the three-dimensional space coordinate system; p is the range of the hyperplane obtained after the n top plate pressure mark data are expanded in the three-dimensional space coordinate system;

[0018] According to the formula:

[0019]

[0020] Determine the jth top plate pressure mark data X( j ) to the shortest distance of the hyperplane obtained after the n top plate pressure mark data are expanded in the three-dimensional space coordinate system;

[0021] And according to the formula:

[0022]

[0023] and:

[0024] Y (j) (W T ·X (j) +p)>0

[0025] Where Y (j) is an intermediate variable, for the top plate pressure marking data X( j ) are assigned values ​​so that different top plate pressure mark data are distributed on both sides of the hyperplane obtained after the n top plate pressure mark data are unfolded in the three-dimensional space coordinate system; a and b respectively represent different clustering levels of the top plate pressure mark data.

[0026] Preferably, in step S103, according to the formula:

[0027] f(X (i) )=W T ·X (j) +p

[0028] Determine the roof pressure data X received by the regional edge gateway within the time period [t3, t4] (i) to the vector value f(X (i) );

[0029] According to the top plate pressure data X (i) to the vector value f(X(i) ) is positive or negative, the top plate pressure data X (i) to classify.

[0030] The embodiment of the present application also provides a coal mining face roof disaster monitoring system based on edge computing, including:

[0031] A data acquisition unit is configured to evenly arrange a plurality of pressure radio frequency sensors on the hydraulic support of the coal mining face to monitor the roof pressure data of the coal mining face and transmit it wirelessly through a built-in LoRa module; and a regional edge gateway is arranged in the air intake lane of the coal mining face to receive the roof pressure data monitored by the pressure radio frequency sensor;

[0032] A hyperplane unit, configured to cause the regional edge gateway to mark the received roof pressure data within the time period [t1, t2], and determine an optimal hyperplane based on the obtained roof pressure mark data;

[0033] A disaster determination unit is configured to determine the vector direction of the roof pressure data within the time period [t3, t4] received by the regional edge gateway to the optimal hyperplane, so as to classify the roof pressure data within the received time period [t3, t4], and determine the roof disaster risk area of ​​the coal mining face during the production process according to the classification result; wherein the time period [t3, t4] is later than the time period [t1, t2].

[0034] Beneficial effects:

[0035] In the method for monitoring roof disasters of coal mining faces based on edge computing provided in an embodiment of the present application, a plurality of pressure RF sensors are evenly arranged on the hydraulic supports of the coal mining faces to monitor the roof pressure data of the coal mining faces and transmit them wirelessly through the built-in LoRa module; at the same time, a regional edge gateway is arranged in the air intake lane of the coal mining faces to receive the roof pressure data monitored by the pressure RF sensor; then, the roof pressure data within the time period [t1, t2] received by the regional edge gateway is marked, and the optimal hyperplane is determined based on the roof pressure marked data obtained by the marking; finally, the vector direction of the roof pressure data within the time period [t3, t4] received by the regional edge gateway to the optimal hyperplane is determined to classify the received roof pressure data within the time period [t3, t4], and the roof disaster risk area of ​​the coal mining faces in the production process is determined based on the classification results.

[0036] Therefore, the collected roof pressure data is transmitted by wireless transmission, and the collected data is analyzed and processed through the working face area edge gateway, so as to obtain the roof activity law of the working face in scientific, real-time and accurate manner, determine the dangerous range of roof disasters, and greatly improve the timeliness of the working face roof disaster analysis; at the same time, the roof pressure data is processed in real time through the regional edge gateway to realize real-time active perception and intelligent analysis, form the best decision, and divide the dangerous range of roof disasters, so as to more effectively prevent and control the working face roof disasters. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The drawings in the specification, which constitute a part of the present application, are used to provide further understanding of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute improper limitations on the present application.

[0038] in:

[0039] Figure 1 A schematic flow chart of a method for monitoring roof disasters in a coal mining face based on edge computing according to some embodiments of the present application;

[0040] Figure 2 A schematic diagram of the arrangement of a roof pressure data acquisition device provided according to some embodiments of the present application;

[0041] Figure 3 A schematic diagram of roof pressure data classification of a coal mining face provided according to some embodiments of the present application;

[0042] Figure 4 A structural schematic diagram of a coal mining face roof disaster monitoring system based on edge computing provided according to some embodiments of the present application. DETAILED DESCRIPTION

[0043] The present application will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments. Each example is provided by way of explanation of the present application and does not limit the present application. In fact, it will be clear to those skilled in the art that modifications and variations may be made in the present application without departing from the scope or spirit of the present application. For example, a feature shown or described as a part of an embodiment may be used in another embodiment to produce yet another embodiment. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the embodiments of the present invention should belong to the scope of protection of the embodiments of the present invention.

[0044] At present, in the monitoring method of roof disasters, before monitoring, it is necessary to evaluate the stability of the tunnel surrounding rock, and determine the monitoring range of the surrounding rock danger area based on the surrounding rock stability evaluation results, and then determine the severe disaster area of ​​the tunnel roof, and arrange monitoring instruments in the severe disaster area for monitoring. However, this method lacks the verification of the reliability of the evaluation indicators and indicator thresholds, and does not take into account the differences in indicator thresholds under different geological conditions, resulting in a large discrepancy between the monitoring results and the actual results, and the monitoring reliability is poor.

[0045] In the monitoring method that collects data through sensors and transmits the data to the ground using switches and monitoring substations, the ground server evaluates the roof status through mathematical models and indicator analysis to give different early warning signals. However, because the data needs to be transmitted from the underground to the ground, the data response time is long, resulting in low timeliness and poor accuracy of the monitoring results.

[0046] Based on this, this application proposes a method for monitoring roof disasters in coal mining working faces based on edge computing. The method uses wireless transmission to collect index data, and analyzes and processes the collected data at the working face data end through the working face area edge gateway, so as to scientifically, real-timely and accurately obtain the roof activity rules of the working face and determine the dangerous range of roof disasters. Figures 1 to 3 As shown, the method for monitoring roof disasters in a coal mining face based on edge computing includes:

[0047] Step S101, evenly arrange multiple pressure RF sensors on the hydraulic support of the coal mining working face to monitor the roof pressure data of the coal mining working face and transmit it wirelessly through the built-in LoRa module; and arrange a regional edge gateway in the air intake lane of the coal mining working face to receive the roof pressure data monitored by the pressure RF sensor.

[0048] In this application, the pressure RF sensors are evenly arranged on the hydraulic supports of the working face to fully cover and monitor the roof pressure of the coal mining working face; wherein, the pressure RF sensors are equipped with a built-in wireless LoRa module for wireless communication to wirelessly transmit the monitored roof pressure data.

[0049] In the air intake tunnel of the coal mining face, regional edge gateways, transformer boxes, mobile substations, etc. are arranged. The regional edge gateways, transformer boxes, and mobile substations are electrically connected to ensure the normal power supply of the regional edge gateways. The roof pressure time monitored by the pressure RF sensor is received by the regional edge gateway, and real-time processing and analysis are performed underground to greatly improve the real-time and effectiveness of data monitoring.

[0050] Step S102: The regional edge gateway marks the received roof pressure data within the time period [t1, t2], and determines the optimal hyperplane according to the obtained roof pressure marked data.

[0051] In this application, the pressure RF sensor monitors the roof pressure of the coal mining face and sends it to the regional edge gateway via wireless transmission. After the regional edge gateway receives the roof pressure data sent by the pressure RF sensor, it divides the received roof pressure data into two parts. The earlier part is used as sample data to seek the optimal hyperplane, and the later part is classified according to the determined optimal hyperplane to determine the roof disaster risk area of ​​the coal mining face during the production process.

[0052] Alternatively, the optimal hyperplane is determined by collecting historical data of roof pressure of the coal mining face; then, the roof pressure data collected in real time by the pressure RF sensor is classified by the optimal hyperplane to determine the roof disaster risk area of ​​the coal mining face during the production process. In other words, the roof pressure data in the time period [t1, t2] can be the earlier part of the roof pressure data collected in real time by the pressure RF sensor, or the historical data of the roof pressure of the coal mining face.

[0053] Here, the received roof pressure data in the time period [t1, t2] are marked based on the clustering algorithm to obtain the roof pressure marked data. Specifically, according to the formula:

[0054]

[0055] Determine the i-th top plate pressure data x received in the time period [t1, t2] i To the predetermined mth cluster center c m Where n is the number of roof pressure data in the time period [t1, t2], n is a positive integer, m = 2. Then, according to the i-th roof pressure data x i To the mth cluster center c m For the i-th top plate pressure data x i Perform clustering level labeling.

[0056] Repeat the cycle to assign the i-th top plate pressure data x i To the nearest cluster center, and update the cluster center cyclically until the cluster center no longer changes significantly, or the number of cyclic iterations reaches the set value. Based on the two cluster centers, the roof pressure data with low values ​​is classified as "normal" level, recorded as "a", and the roof pressure data with high values ​​is classified as "dangerous" level, recorded as "b".

[0057] Next, the marked roof pressure data is expanded in a three-dimensional space coordinate system with the working face advancement distance, hydraulic support number, and pressure value as characteristic indicators. According to the distribution characteristics of the data, there will be a hyperplane in the space to separate the high pressure value and low pressure value data. In order to accurately classify these two types of data, it is necessary to ensure that the distance from the points on the boundary of the two types of data to the hyperplane is the largest. Therefore, based on the preset objective function:

[0058] max margin(W, p)

[0059] Determine the optimal hyperplane. In the formula, W is the direction vector of the hyperplane obtained after the n top plate pressure marker data are expanded in the three-dimensional space coordinate system; p is the range of the hyperplane obtained after the n top plate pressure marker data are expanded in the three-dimensional space coordinate system.

[0060] Here, the objective function must satisfy two constraints: one is that the distance from the point in the data set to the hyperplane is the shortest; the other is to ensure that data with different labels are distributed twice in the hyperplane. Based on this, in this application, according to the formula:

[0061]

[0062] Determine the jth top plate pressure mark data X( j ) to the shortest distance of the hyperplane obtained after the n top plate pressure mark data are expanded in the three-dimensional space coordinate system;

[0063] And according to the formula:

[0064]

[0065] and:

[0066] Y (j) (W T ·X (j) +p)>0

[0067] In the formula, a and b represent different clustering levels of roof pressure label data, respectively. "a" means that the roof pressure data is classified as "normal", that is, the roof pressure at the corresponding point is in a safe range; "b" means that the roof pressure data is classified as "dangerous", that is, the roof pressure at the corresponding point is in a dangerous range. T Represents the transpose of the direction vector W of the hyperplane obtained after the n top plate pressure mark data are expanded in the three-dimensional space coordinate system.

[0068] Y (j) is the intermediate variable, for the top plate pressure label data X (j)Assign values ​​so that different roof pressure mark data are distributed on both sides of the hyperplane obtained after the n roof pressure mark data are expanded in the three-dimensional space coordinate system. (j) Assigning values ​​to different types of data can enable different types of data to be accurately classified to both sides of the hyperplane.

[0069] Step S103, determine the vector direction of the roof pressure data within the time period [t3, t4] received by the regional edge gateway to the optimal hyperplane, so as to classify the received roof pressure data within the time period [t3, t4], and determine the roof disaster risk area of ​​the coal mining working face in the production process according to the classification results.

[0070] In the present application, the earlier part of the roof pressure data collected in real time by the pressure radio frequency sensor, that is, the roof pressure data in the time period [t1, t2], or the historical data of the roof pressure of the coal mining face is used as sample data to determine the optimal hyperplane, and the roof pressure data collected in real time by the pressure radio frequency sensor or the later part of the roof pressure data collected in real time by the pressure radio frequency sensor, that is, the roof pressure data in the time period [t3, t4] are classified. In other words, the roof pressure data in the time period [t3, t4] can be all the roof pressure data collected in real time by the pressure radio frequency sensor, or it can be the data in the time period [t3, t4] that is later than the time period [t1, t2].

[0071] The optimal hyperplane is determined by the roof pressure data of the time period [t1, t2], and then the roof pressure data X of the time period [t3, t4] is (i) Expand in the three-dimensional space coordinate system, and then convert the top plate pressure data X (i) Substitute the determined optimal hyperplane equation:

[0072] W T ·X (j) +p=0

[0073] By calculating:

[0074] f(X (i) )=W T ·X (j) +p

[0075] Determine the roof pressure data X received by the regional edge gateway in the time period [t3, t4] (i) The vector value f(X (i) ).

[0076] Then, according to the top plate pressure data X (i) The vector value f(X (i)) is positive or negative, and has an influence on the top plate pressure data X (i) Specifically, when f(X (i) )>0, the top plate pressure data X (i) Classified as "dangerous" when (X (i) )<0, the top plate pressure data X (i) Classified as "Normal".

[0077] Finally, according to the classification results of the roof pressure data in the time period [t3, t4], the roof disaster danger zone of the coal mining face in the production process is determined, that is, the area corresponding to the roof pressure data classified as "dangerous" in the time period [t3, t4] is the roof disaster danger zone of the coal mining face in the production process.

[0078] By dividing the roof disaster risk area of ​​the coal mining face during the production process, the step distance and duration of the pressure on the coal mining face can be reflected. The pressure on the coal mining face usually changes periodically, which can better reflect the law of roof activity. In this way, the roof pressure data is processed in real time through the regional edge gateway, and real-time active perception and intelligent analysis are achieved to form the best decision, divide the roof disaster risk range, and thus more effectively prevent and control the roof disaster on the working face.

[0079] The embodiment of the present application also provides a coal mining face roof disaster monitoring system based on edge computing, such as Figure 3 As shown, the monitoring system includes:

[0080] The data acquisition unit 301 is configured to evenly arrange a plurality of pressure RF sensors on the hydraulic support of the coal mining face to monitor the roof pressure data of the coal mining face and transmit it wirelessly through the built-in LoRa module; and to arrange a regional edge gateway in the air intake lane of the coal mining face to receive the roof pressure data monitored by the pressure RF sensor;

[0081] The hyperplane unit 302 is configured as a regional edge gateway to mark the received roof pressure data in the time period [t1, t2] and determine the optimal hyperplane according to the obtained roof pressure data;

[0082] The disaster determination unit 303 is configured to determine the vector direction of the roof pressure data within the time period [t3, t4] received by the regional edge gateway to the optimal hyperplane, so as to classify the received roof pressure data within the time period [t3, t4], and determine the roof disaster risk area of ​​the coal mining working face in the production process according to the classification result; wherein, the time period [t3, t4] is later than the time period [t1, t2].

[0083] The edge computing-based coal mining face roof disaster monitoring system provided in the embodiment of the present application can implement the steps and processes of any of the above-mentioned edge computing-based coal mining face roof disaster monitoring methods, and achieve the same technical effects, which will not be repeated here one by one.

[0084] In the description of the present invention, the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" etc. mean that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0085] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for monitoring roof disasters in a coal mining face based on edge computing, characterized in that: include: Step S101, evenly arranging a plurality of pressure radio frequency sensors on the hydraulic support of the coal mining working face to monitor the roof pressure data of the coal mining working face and wirelessly transmitting the data through the built-in LoRa module; and arranging a regional edge gateway in the air intake lane of the coal mining working face to receive the roof pressure data monitored by the pressure radio frequency sensor; Step S102, the regional edge gateway marks the received roof pressure data in the time period [t1, t2], and determines the optimal hyperplane according to the obtained roof pressure marked data; Step S103, determine the vector direction of the roof pressure data in the time period [t3, t4] received by the regional edge gateway to the optimal hyperplane, so as to classify the roof pressure data in the received time period [t3, t4], and determine the roof disaster risk area of ​​the coal mining working face in the production process according to the classification result; wherein the time period [t3, t4] is later than the time period [t1, t2].

2. The method for monitoring roof disasters in a coal mining face based on edge computing according to claim 1, characterized in that: In step S102, The received roof pressure data in the time period [t1, t2] is labeled based on a clustering algorithm to obtain the roof pressure labeled data, and the optimal hyperplane is determined according to the roof pressure labeled data.

3. The method for monitoring coal mining face roof disasters based on edge computing according to claim 2, characterized in that: The step of marking the received top plate pressure data within the time period [t1, t2] based on a clustering algorithm includes: According to the formula: Determine the top plate pressure data x received in the time period [t1, t2] of the i i To the predetermined mth cluster center c m Wherein, n is the number of the top plate pressure data in the time period [t1, t2]; n is a positive integer; m=2; According to the top plate pressure data x i To the mth cluster center c m The distance d(i,m) of the top plate pressure data x i Perform clustering level labeling.

4. The method for monitoring coal mining face roof disasters based on edge computing according to claim 2, characterized in that: The determining the optimal hyperplane according to the top plate pressure mark data comprises: The roof pressure marking data is expanded in a three-dimensional space coordinate system with the working face advancement distance, hydraulic support number and pressure value as characteristic indicators, and the optimal hyperplane is determined based on a preset objective function.

5. The method for monitoring roof disasters in a coal mining face based on edge computing according to claim 4 is characterized in that: The determining the optimal hyperplane based on a preset objective function includes: Based on the objective function: max margin(W,p) determining the optimal hyperplane; Wherein, W is the direction vector of the hyperplane obtained after the n top plate pressure mark data are expanded in the three-dimensional space coordinate system; p is the range of the hyperplane obtained after the n top plate pressure mark data are expanded in the three-dimensional space coordinate system; According to the formula: Determine the jth top plate pressure mark data X (j) The shortest distance to the hyperplane obtained after the n top plate pressure mark data are expanded in the three-dimensional space coordinate system; And according to the formula: and: Y (j) (W T ·X (j) +p)>0 Where Y (j) is an intermediate variable, and for the jth top plate pressure label data X (j) Assign values ​​so that different roof pressure mark data are distributed on both sides of a hyperplane obtained after the n roof pressure mark data are unfolded in a three-dimensional space coordinate system; a and b respectively represent different clustering levels of the roof pressure mark data.

6. The method for monitoring coal mining face roof disasters based on edge computing according to claim 5, characterized in that: In step S103, According to the formula: f(X (i) )=W T ·X (j) +p Determine the roof pressure data X received by the regional edge gateway within the time period [t3, t4] (i) to the vector value f(X (i) ); According to the top plate pressure data X (i) to the vector value f(X (i) ) is positive or negative, the top plate pressure data X (i) to classify.

7. A coal mining face roof disaster monitoring system based on edge computing, characterized in that: include: A data acquisition unit is configured to evenly arrange a plurality of pressure radio frequency sensors on the hydraulic support of the coal mining face to monitor the roof pressure data of the coal mining face and transmit it wirelessly through a built-in LoRa module; and a regional edge gateway is arranged in the air intake lane of the coal mining face to receive the roof pressure data monitored by the pressure radio frequency sensor; A hyperplane unit, configured to cause the regional edge gateway to mark the received roof pressure data in the time period [t1, t2], and determine an optimal hyperplane according to the obtained roof pressure mark data; A disaster determination unit is configured to determine the vector direction of the roof pressure data within the time period [t3, t4] received by the regional edge gateway to the optimal hyperplane, so as to classify the received roof pressure data within the time period [t3, t4], and determine the roof disaster risk area of ​​the coal mining face during the production process according to the classification result; wherein the time period [t3, t4] is later than the time period [t1, t2].

Images