Coal mine methane leakage early warning method and system based on wireless sensor network

Abstract

The invention discloses a coal mine methane leakage early warning method and system based on a wireless sensor network, and relates to the technical field of leakage early warning. The method comprises the following steps: deploying a sparse sensor network in a coal mine area, and constructing a coal mine cloud picture; real-time data of a coal mine area are collected and transmitted back to the central control processor; if the real-time data represents that no leakage exists, but the real-time data is in a metastable critical state, carrying out directional sensing management by calling a mobile sensing component in the sparse sensing network; if the real-time data represents that leakage exists, a leakage verification assembly is triggered, the data are synchronized to a coal mine cloud picture, cloud picture network screening and superposition state definition based on leakage events are carried out, and leakage data are determined and leakage early warning management is carried out based on superposition state collapse of candidate leakage paths by introducing a walking operator. The technical problem that in the prior art, coal mine methane leakage monitoring depends on a fixed sensing point, and the leakage state cannot be dynamically tracked is solved, and the technical effect of coal mine methane leakage dynamic positioning is achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of leakage early warning, in particular to a coal mine methane leakage early warning method and system based on a wireless sensor network. BACKGROUND

[0002] Methane gas in a coal mine operation environment has highly flammable and explosive characteristics, and its leakage monitoring and early warning is an important link to ensure the safety of mine production. The existing methane leakage monitoring system mostly adopts a fixed sensor point layout method, and periodically samples and uploads the methane concentration through wired or wireless methods. However, such a system has obvious deficiencies: on the one hand, the density of the sensor point layout is limited, the monitoring spatial resolution is low, and it cannot comprehensively cover complex roadways and ventilation blind areas; on the other hand, fixed point monitoring cannot reflect the spatio-temporal dynamic changes of the methane concentration, and it is prone to detection delay or false reporting in the early or diffusion stage of leakage. In addition, the existing monitoring scheme is mostly based on threshold determination, and lacks the ability to identify sub-stable critical states, making it difficult to discover potential leakage risks in a timely manner. SUMMARY

[0003] The present application provides a coal mine methane leakage early warning method and system based on a wireless sensor network, which solves the technical problem that the existing coal mine methane leakage monitoring relies on fixed sensor points and cannot dynamically track the leakage state.

[0004] In a first aspect, the present application provides a coal mine methane leakage early warning method based on a wireless sensor network, which comprises: A sparse sensor network is deployed in a coal mine area to construct a coal mine cloud map; the sparse sensor network performs real-time data collection in the coal mine area and returns to a central control processor; if the real-time data indicates that there is no leakage but is in a sub-stable critical state, a mobile sensor component in the sparse sensor network is called to perform directional sensing management, wherein the leakage is determined based on the exceeding of the methane concentration of any sensor data; if the real-time data indicates that there is leakage, a leakage determination component is triggered, data is synchronized to the coal mine cloud map, and cloud map network screening and superposition state definition based on the leakage event are performed, a walking operator is introduced to determine the leakage data based on the superposition state collapse of the candidate leakage path, and leakage early warning management is performed, wherein the leakage data includes the leakage source position and the target leakage path.

[0005] In a second aspect, the present application provides a coal mine methane leakage early warning system based on a wireless sensor network, which comprises: The cloud map construction module: a sparse sensor network is deployed in a coal mine area to construct a coal mine cloud map; the data acquisition module: the sparse sensor network performs real-time data acquisition of the coal mine area and returns to the central control processor; the sensing management module: if the real-time data indicates that there is no leakage but is in a metastable critical state, a mobile sensing component in the sparse sensor network is called to perform directional sensing management, wherein the leakage is determined according to the exceeding of the methane concentration of any sensing data; the early warning module: if the real-time data indicates that there is leakage, a leakage detection component is triggered, data is synchronized to the coal mine cloud map, and cloud map network screening and superposition state definition based on a leakage event are performed, a walking operator is introduced to collapse the superposition state based on a candidate leakage path, leakage data including a leakage source position and a target leakage path are determined, and leakage early warning management is performed.

[0006] One or more technical solutions provided in the present application have at least the following technical effects or advantages: A sparse sensor network is deployed in a coal mine area to construct a coal mine cloud map; the sparse sensor network performs real-time data acquisition of the coal mine area and returns to the central control processor. If the real-time data indicates that there is no leakage but is in a metastable critical state, a mobile sensing component in the sparse sensor network is called to perform directional sensing management, wherein the leakage is determined according to the exceeding of the methane concentration of any sensing data. If the real-time data indicates that there is leakage, a leakage detection component is triggered, data is synchronized to the coal mine cloud map, and cloud map network screening and superposition state definition based on a leakage event are performed, a walking operator is introduced to collapse the superposition state based on a candidate leakage path, leakage data including a leakage source position and a target leakage path are determined, and leakage early warning management is performed. The technical problem that the coal mine methane leakage monitoring in the prior art relies on fixed sensing points and cannot dynamically track the leakage state is solved, and the technical effect of dynamic positioning of coal mine methane leakage is achieved. BRIEF DESCRIPTION OF DRAWINGS

[0007] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0008] Figure 1 A flowchart of a coal mine methane leakage early warning method based on a wireless sensor network is provided for the embodiments of the present application. Figure 2 A structure diagram of a coal mine methane leakage early warning system based on a wireless sensor network is provided for the embodiments of the present application.

[0009] Explanation of reference signs: cloud map construction module 11, data acquisition module 12, sensor management module 13, early warning module 14. DETAILED DESCRIPTION

[0010] In order to further illustrate the technical means and effects adopted by the present application to achieve the predetermined object of the application, the specific embodiments, structures, features and effects according to the present application are described in detail below in combination with the drawings and preferred embodiments.

[0011] In one embodiment, as shown in the accompanying drawings, the present application provides a coal mine methane leakage early warning method based on a wireless sensor network, wherein the method comprises: Figure 1 deploying a sparse sensor network in a coal mine area to construct a coal mine cloud map.

[0012] In the present application, according to the structure of a coal mine underground roadway, the ventilation route and the distribution characteristics of historical leakage events, the sensor deployment area and the sparse sampling interval are determined; according to a preset spatial layering rule, the coal mine area is divided into a plurality of sensing sub-areas, a plurality of fixed sensing nodes and movable sensing nodes are arranged in each sub-area, and a sparse sensor network is formed. The sensing nodes at least include a methane sensing unit for detecting methane concentration, an environmental sensing unit for collecting wind speed and wind direction, and an inertial measurement unit for positioning.

[0013] After the node deployment is completed, based on the collected spatial coordinates of the sensing nodes, the positions of the ventilation facilities, the potential leakage points and the monitoring blind area information, the nodes are numbered and mapped, and the directed edges are defined according to the wind flow direction and the roadway connection relationship; through the association relationship between the nodes and the edges, a graph model describing the spatial structure of the coal mine and the gas flow characteristics is constructed. Based on the graph model, a coal mine cloud map data structure is established, the coal mine cloud map is a multi-dimensional fusion model, including a node attribute matrix, an edge weight matrix and a real-time sensing data matrix, and is used to represent the dynamic gas distribution state of the coal mine area.

[0014] Further, deploying a sparse sensor network in a coal mine area to construct a coal mine cloud map comprises: defining graph nodes based on the sensing nodes of the sparse sensor network, the positions of the ventilation facilities, the potential leakage points and the monitoring blind area, wherein the sparse sensor network at least includes a methane sensing type and an environmental sensing type; defining directed edges according to the wind flow direction and the roadway connection relationship; and constructing the coal mine cloud map according to the graph nodes and the directed edges.

[0015] ​According to the spatial topology structure, roadway trend and ventilation network layout of the underground coal mine, the laying area and node density of the sparse sensor network are determined. Key spatial elements of the coal mine area are mapped into graph nodes, which include: a plurality of sensing nodes based on the sparse sensor network, each sensing node carrying a unique node number and three-dimensional spatial coordinate information, for collecting real-time environmental parameters such as methane concentration, wind speed, temperature, humidity and air pressure; ventilation facility location nodes for reflecting the spatial position and ventilation direction of ventilation equipment such as fans, air doors, air ducts and the like in the network; potential leakage point nodes for identifying areas that may have a risk of methane seepage, such as gas extraction boreholes, mining working faces and sealed space boundaries; and monitoring blind area nodes for supplementing the monitoring gap areas caused by terrain obstruction or ventilation dead angles.

[0016] The sparse sensor network at least includes two categories of methane sensing nodes and environmental sensing nodes, wherein the methane sensing nodes are used for real-time detection of methane concentration distribution, and the environmental sensing nodes are used for providing external environmental parameters affecting methane diffusion. Based on the air flow direction in the coal mine ventilation system, the geometric connection relationship of the roadway and the gas flow constraint conditions, directed edges are established between the above nodes to represent the gas transport direction and ventilation connectivity. The weight of the directed edge is valued according to the wind speed, the roadway cross-sectional size, the local resistance coefficient and other dynamic parameters, for representing the transmission strength of gas flow. According to the set of graph nodes and directed edges, a coal mine cloud chart is constructed, which is a high-dimensional graph model describing the dynamic distribution state of gas in the coal mine area.

[0017] Further, after the coal mine cloud chart is constructed, the following steps are included: uploading coal mine condition variables, wherein the coal mine condition variables at least include working face opening and ventilation route change; and updating the coal mine cloud chart according to the coal mine condition variables.

[0018] Dynamic change information in the coal mine operating environment is uploaded to the central control processor as an input parameter to realize adaptive updating of the coal mine cloud chart. Specifically, coal mine condition variables are collected and uploaded, which are used to represent structural changes in the coal mine production environment or ventilation system; working face opening information: when a new mining working face or roadway is opened, the spatial coordinates, length, cross-sectional size and ventilation connection relationship of the newly opened area are automatically recorded through the mining equipment control system and the geographic information subsystem, and the above information is uploaded to the central control processor; ventilation route change information: when the air flow path is changed due to construction adjustment, local sealing, air door opening and closing or air duct extension of the coal mine ventilation system, the new ventilation node, air flow direction and flow change parameter are recorded through the ventilation monitoring device or manual input method.

[0019] After receiving the coal mine condition variables, the central control processor automatically updates the structure of the coal mine cloud map: first, identify the newly added or failed nodes and directed edges involved in the condition variables, and adjust the node set and edge set accordingly; second, according to the new ventilation connection relationship and the wind flow direction, recalculate the weight parameters of each directed edge and correct the transmission path topology between nodes; finally, manage the version and time sequence of the updated coal mine cloud map, and generate the corresponding cloud map update timestamp to ensure that the cloud map data called by the subsequent leakage detection and early warning algorithm is consistent with the current state of the coal mine site.

[0020] The sparse sensing network performs real-time data collection in the coal mine area and returns to the central control processor.

[0021] Each type of sensing node in the sparse sensing network collects environmental parameters such as methane concentration, wind speed, temperature, humidity, and air pressure in the coal mine area according to a preset sampling period. Each sensing node has a micro-processing unit built-in for local preprocessing operations such as data sampling, noise filtering, and threshold judgment. The collected multi-dimensional environmental data is identified by timestamp and node number to form a standardized data frame.

[0022] At the data transmission level, each sensing node constructs a low-power self-organizing network through a wireless communication protocol and uses a multi-hop relay method to return the collected data to the central control processor.

[0023] After receiving real-time data from each node, the central control processor first matches and corrects the data according to the node number and spatial coordinate information; second, it performs time synchronization and abnormality rejection algorithms to interpolate and correct delayed or lost data packets; finally, it updates the processed data to the node attribute matrix in the coal mine cloud map to form a dynamic visual distribution map reflecting the current methane concentration and environmental state in the mine area.

[0024] If the real-time data indicates no leakage but is in a metastable critical state, directed sensing management is performed by calling the mobile sensing components in the sparse sensing network, where the methane concentration exceeding the limit of any sensing data is used to determine leakage.

[0025] When analyzing the real-time data collected by the sparse sensing network, the central control processor determines whether the current coal mine area is in a metastable critical state based on the time series change rate of methane concentration, spatial concentration gradient, and ventilation flow field disturbance characteristics. The metastable critical state refers to a state where the methane concentration does not exceed the safety threshold, but the concentration fluctuation amplitude, local gradient mutation rate, or wind flow disturbance intensity exceeds the preset safety margin, indicating a potential leakage risk in the coal mine area.

[0026] When the metastable critical state is detected, the central control processor automatically generates dynamic sensing instructions to call mobile sensing components (such as track mobile sensing vehicles, unmanned aerial vehicles or tracked sampling robots) in the sparse sensing network to perform directional sensing tasks. The mobile sensing components move along the airflow path of the high-risk area according to the node distribution and ventilation direction information in the coal mine cloud map for mobile sampling; the moving path is dynamically adjusted according to the concentration gradient direction, and the space area with a larger concentration change rate is preferentially covered to improve the capture probability of potential leakage points.

[0027] During the directional sensing process, the mobile sensing components real-time return the measurement data such as methane concentration, wind speed and temperature, and the central control processor performs spatial fusion calculation on these data and the data of the sparse fixed nodes to form a local high-precision concentration field distribution.

[0028] If the methane concentration value detected by any sensing node exceeds the safety threshold, it is defined as a methane concentration out-of-limit event, the system immediately determines that there is a leakage, and enters the leakage determination stage.

[0029] Further, the directional sensing management by calling the mobile sensing components in the sparse sensing network includes: defining a macroscopic behavior variable, wherein the macroscopic behavior variable is defined according to the overall behavior description of the coal mine area, and at least includes concentration field information entropy, wind flow vortex intensity and sensing network effective information flow; generating a dynamic sensing instruction according to the macroscopic behavior variable, wherein the generation condition of the dynamic sensing instruction is the probability of the existence of a metastable critical state; calling the mobile sensing components in the sparse sensing network according to the dynamic sensing instruction to perform directional mobile sensing management; wherein the coal mine cloud map is synchronously updated with the change of the position of the mobile sensing components.

[0030] The macroscopic behavior variable is used to describe the overall operation state and gas flow characteristics of the coal mine area, and comprehensively reflects the methane diffusion dynamics and the stability of the ventilation flow field. Specifically, the macroscopic behavior variable at least includes: concentration field information entropy: the information entropy of the whole area methane concentration distribution is calculated to evaluate the uncertainty degree of the concentration distribution, and the higher the information entropy, the more discrete the concentration distribution and the lower the system stability; wind flow vortex intensity: the local vorticity is calculated based on the velocity vector data of the ventilation flow field, which is used to reflect the potential stagnation and backflow phenomenon in gas flow; sensing network effective information flow: the effective data amount uploaded by the sensing nodes and the communication delay between nodes per unit time are taken as indicators to represent the activity and data coverage ability of the current sparse sensing network.

[0031] When the concentration field information entropy increase rate exceeds the preset threshold, the wind vortex intensity significantly increases, or the effective information flow of the sensing network decreases, the central control processor determines that the system has a sub-stable critical state; in this case, a dynamic sensing instruction is generated, the spatial weight distribution of the high-risk area is calculated, the area range that needs to be intensively monitored is determined, and the corresponding directional sensing path and task parameters are generated. The mobile sensing component executes mobile monitoring according to the path parameters of the dynamic sensing instruction, and the mobile trajectory can be along the main air duct, the branch lane or the working face boundary. The mobile sensing component dynamically adjusts the sampling step and speed according to the real-time concentration change rate during movement, realizing directional mobile sensing management.

[0032] During the execution of the task by the mobile sensing component, the coal mine cloud map is updated synchronously with the change of the position: the central control processor receives the position information and measurement data of the mobile sensing node in real time, maps it to the corresponding spatial node in the coal mine cloud map, and incrementally updates the concentration field, flow field and information flow matrix of the local node.

[0033] If the real-time data represents a leak, the leak detection component is triggered, the data is synchronized to the coal mine cloud map and subjected to cloud network screening and superposition state definition based on the leakage event, the leakage data containing the leakage source position and the target leakage path are determined by introducing a walk operator to collapse the superposition state based on the candidate leakage path, and leakage warning management is performed.

[0034] When the central control processor detects that the methane concentration of any node exceeds the safety threshold or the concentration growth rate shows an exponential upward trend in a plurality of consecutive sampling periods according to the real-time data of the sensing network, it is determined that there is a leakage risk in the coal mine area, and the leak detection component is started.

[0035] The leak detection component first synchronizes the time series data of the abnormal node and its surrounding neighborhood nodes to the coal mine cloud map, and generates a leakage event identification layer accordingly. Subsequently, the system performs network screening and superposition state definition based on the leakage event in the coal mine cloud map. Specifically, the central control processor screens out the air flow path subgraph connected to the node from the graph structure of the coal mine cloud map according to the position of the abnormal node and the ventilation direction, forming a leakage event network; in the leakage event network, the state (including concentration, wind speed and air pressure change) of each node is initialized as a probability superposition state to represent the multiple candidate diffusion paths through which the leaked gas may pass. The probability distribution of the superposition state is determined by the edge weight (wind speed direction and strength) between nodes, the concentration gradient difference and the time correlation, so that the coal mine cloud map can quantitatively represent the uncertainty of leakage diffusion.

[0036] The system introduces a walking operator to perform dynamic convergence analysis on the candidate paths. The walking operator is constructed based on the quantum walking idea and performs multi-step random walking operation in the leakage event network under the constraint of gas diffusion direction, wind flow field and concentration gradient. Each step of walking corresponds to a diffusion simulation of the candidate path, and the probability superposition state of the leakage path is formed by superimposing multiple steps. After completing the multi-step walking, the system performs collapse operation on the superposition state, that is, the probability amplitude of each candidate path is normalized and selected to make the path with the highest compliance (i.e., the highest consistency of concentration gradient matching degree and time phase) collapse into the target path, and obtain a unique target leakage path. At the same time, the steady-state probability density center of the starting point of the path after the collapse of the superposition state is calculated as the location of the leakage source.

[0037] The system performs leakage early warning management according to the leakage data: labels the leakage source and path information in the unified three-dimensional coordinate system of the coal mine cloud map, and automatically generates an alarm identifier; the central control processor pushes the leakage level, diffusion trend and risk area range to the upper safety monitoring system or personnel terminal, realizes real-time early warning and emergency disposal linkage of leakage.

[0038] Further, determining the leakage data comprises: Based on the wind flow field of the coal mine regional ventilation network, the real-time diffusion condition is determined; the real-time concentration signals of each methane sensing network are read, the leakage event network is selected in the coal mine cloud map, and each network node is initialized with a superposition state as a leakage event network; based on the leakage event network and the real-time diffusion condition, the leakage source is located.

[0039] The central control processor constructs a wind flow field model according to the directed edge weight, wind speed vector and air pressure distribution information of the ventilation system in the coal mine cloud map. The wind flow field model obtains real-time gas diffusion condition parameters, including local flow direction, flow size, pressure gradient and vortex distribution characteristics, by numerically solving the steady-state equation and local disturbance equation of the ventilation network.

[0040] The central control processor calls the real-time concentration data of all methane sensing nodes in the sparse sensing network and calculates the rate of change of time; when an abnormal concentration node is detected, a leakage event network is constructed by selecting a number of connected nodes upstream and downstream of the node in the coal mine cloud map according to the ventilation flow direction.

[0041] In the leakage event network, the state of each node is initialized as a superposition state: the concentration change rate of the node, the concentration difference of the adjacent node and the time phase difference are quantified as a superposition state probability amplitude to represent the participation probability of the node in the potential leakage path. The superposition state definition can reflect the uncertainty and multi-channel propagation characteristics of the leakage path.

[0042] The central control processor calls a walk operator or a random walk algorithm on the leakage event network, iteratively propagates the probability transmission between nodes, and simulates the diffusion process of the gas along the wind direction. During the propagation process, the superposition state probability of the node is continuously corrected according to the real-time diffusion condition, so as to conform to the actual wind flow potential field constraint; with the increase of the iteration number, the probability distribution graph of the leakage path is gradually formed. The most possible position coordinates of the leakage source are determined by calculating the steady-state center or the maximum amplitude point of the probability distribution; at the same time, the node sequence along the maximum probability gradient direction is extracted as the target leakage path.

[0043] Further, based on the leakage event network and the real-time diffusion condition, the leakage source positioning includes: The walk operator is introduced to explore the leakage path in the leakage event network under the constraint of the real-time diffusion condition, and a candidate leakage path set is determined; the superposition state probability collapse processing is performed on the candidate leakage path set to determine the leakage source position.

[0044] The central control processor defines a walk operator for path exploration in the topological structure of the leakage event network, which takes the node concentration gradient, the wind speed vector and the diffusion potential energy difference as the main weight factors, and is used to describe the probability of gas particles jumping from one node to an adjacent node. Specifically, the walk operator is composed of a transition probability matrix, and the matrix element represents the transition probability from node i to node j, which is calculated by comprehensively considering the wind flow direction coefficient , the concentration gradient amplitude and the local ventilation resistance coefficient , formula: wherein, the function f is used to normalize these parameters, so that the transition probability of each node is 1.

[0045] The walk operator simulates the propagation behavior of the gas in the ventilation network through multi-step random walk, and gradually forms a candidate leakage path set. The formation probability of each path is related to the multiplication value of the superposition state amplitude of each node on the path and the transition weight, which reflects the possibility of the path becoming an actual leakage path. Through multiple rounds of walk and probability superposition, the system generates a set of candidate leakage paths in the leakage event network, and each path includes node sequence, path total weight and time phase information.

[0046] The central control processor normalizes and aggregates the paths based on the probability amplitude distribution of all candidate leakage paths, and performs interference analysis according to the phase correlation between the paths: when the time phases of two paths are consistent and the concentration change directions are the same, a constructive interference operation is performed to enhance the probability amplitude of the path; when the path directions or concentration gradient trends are opposite, a destructive interference operation is performed to weaken the probability amplitude. After multiple rounds of interference superposition, the probability distribution of the candidate path set gradually converges, and finally forms several high-probability aggregation areas in the probability space. The system solves the spatial coordinates of the aggregation area center, takes the area with the maximum steady-state probability density as the leakage source position. At the same time, the corresponding path is identified as the target leakage path, and the leakage source node and path data are written into the leakage information layer of the coal mine cloud map.

[0047] Further, before performing the superposition state probability collapse processing on the candidate leakage path set, the method comprises: According to the real-time concentration signal, a source leakage mode is defined; if the first candidate leakage path does not match the source leakage mode, a path probability decay processing based on destructive interference is performed, wherein the first candidate leakage path is any item in the candidate leakage path set, and the decay scale is positively correlated with the degree of inconsistency; if the first candidate leakage path is highly consistent with the source leakage mode, a path probability amplification processing based on constructive interference is performed, wherein the amplification scale is positively correlated with the degree of consistency.

[0048] The source leakage mode is used to represent the spatio-temporal distribution state of the methane gas in the coal mine area under the current unknown leakage source condition. The central control processor constructs a concentration distribution matrix according to the real-time concentration signal and its time rate of change of each sensing node, in combination with the direction of the ventilation air flow, the air pressure difference and the node spatial distance. By performing gradient analysis and time series filtering on the matrix, the concentration rising trend area, the stable area and the falling area are extracted, and a concentration field equipotential surface distribution map is generated, thereby forming the source leakage mode. The source leakage mode is a snapshot of the gas concentration field observed by the current system, which is used to depict the diffusion trend that may be caused by the potential leakage source.

[0049] The central control processor sequentially calculates the spatial similarity and the time correlation of each candidate leakage path with the source leakage mode to obtain a consistency parameter. The calculation of the consistency can comprehensively consider the average deviation, gradient direction difference and phase difference of the path node concentration distribution, and form a unified quantitative index by weighted fusion, with a value range of [0, 1], and the larger the value, the higher the consistency of the path with the source leakage mode.

[0050] When the first candidate leakage path does not match the source leakage mode, i.e. the consistency is less than a set threshold, the system performs path probability decay processing under destructive interference.

[0051] At this time, the path probability amplitude is attenuated according to the following relationship: wherein, is a decay coefficient, the decay scale is positively correlated with the degree of inconsistency .

[0052] When the first candidate leakage path is highly consistent with the source leakage pattern, that is, the degree of consistency is greater than or equal to a set threshold, the system performs path probability amplification processing under constructive interference. At this time, the path probability amplitude is enhanced according to the following relationship: wherein, is an amplification coefficient, the amplification scale is positively correlated with the degree of consistency .

[0053] After the above interference correction, the probability distribution of each candidate path is dynamically normalized and adjusted once to form a new path probability superposition state, which provides a high confidence input for subsequent superposition state collapse and leakage source determination.

[0054] Further, the superposition state probability collapse processing is performed on the candidate leakage path set to determine the leakage source position, including: According to the source leakage pattern, a network node potential field vector based on the leakage event network is determined to generate a source leakage potential field; based on the source leakage potential field, a first comparison based on the potential field trend is performed on the first candidate leakage path, and a second comparison based on the potential field vector is performed, and the degree of consistency is analyzed jointly; wherein, through the path leakage source position of the target leakage path under extremely high probability collapse of the candidate leakage path set, the leakage source position is taken as the leakage source position.

[0055] According to the source leakage mode, a network node potential field vector based on a leakage event network is determined, wherein an overrun value of each node at a baseline concentration (i.e. a difference between a current concentration of the node and a safety concentration baseline) is taken as a potential field vector amplitude, and a vector direction is defined according to a ventilation flow direction and a concentration gradient direction; a potential field difference ratio (i.e. a ratio of an overrun value difference to a spatial distance between nodes) between any adjacent nodes is taken as a potential field trend, thereby generating a source leakage potential field representing diffusion of methane gas from a high potential area to a low potential area at the current time. Subsequently, based on the source leakage potential field, a set of candidate leakage paths is subjected to a conformity analysis: a first comparison based on the potential field trend is performed on a first candidate leakage path, an average of angles between path node directions and a main gradient direction of the potential field is calculated to determine diffusion trend consistency; and a second comparison based on the potential field vector is performed, a correlation between a potential field modulus sequence of the path nodes and a concentration change rate sequence is calculated to determine energy distribution consistency; a comprehensive conformity parameter is obtained by weighted fusion of the two comparison results, which is used to correct an overlap state probability amplitude of the path. Finally, the system performs overlap state collapse under an extremely high probability on the set of candidate leakage paths after correction, when a maximum value of the path probability distribution reaches a set convergence threshold, a path with the highest probability is determined as a target leakage path, and a potential field energy center (i.e. a node with the most significant potential field gradient change) in the path is taken as a leakage source position; the obtained leakage source position and target path information are written into a coal mine cloud map for three-dimensional visualization and subsequent leakage warning management.

[0056] Further, the leakage warning management includes: A unified three-dimensional coordinate system of the coal mine area is established; leakage identification information is generated based on the leakage data and written into the unified three-dimensional coordinate system as alarm data; and leakage warning management based on the alarm data is performed.

[0057] Based on the spatial data structure of the coal mine geographic information system and the coal mine cloud map, a three-dimensional coordinate system corresponding to the actual mine area topology is constructed. The coordinate system takes a mine area surface reference point as the origin, takes a roadway direction as the X axis, takes a ventilation flow direction as the Y axis, and takes a shaft depth direction as the Z axis, forming a quantifiable spatial positioning framework. All sensing nodes, ventilation facilities, mining working faces, and equipment positions are written into the three-dimensional coordinate system through coordinate mapping, so that the system can accurately position and visually display the leakage source and path.

[0058] The central control processor generates a set of leakage identification information according to the leakage source position contained in the leakage data and the target leakage path information. The leakage identification information at least includes: leakage source spatial coordinates, leakage path node sequence, leakage level (determined by concentration value and diffusion rate), diffusion direction vector and risk radius. After embedding the leakage identification information in a three-dimensional coordinate system, the system automatically draws the leakage source point, path line and diffusion area surface, identifies different leakage levels and risk ranges through color gradient or transparency, and forms an intuitive three-dimensional alarm data model.

[0059] When the leakage level is in the low risk interval, the system only prompts on the monitoring interface and starts the local data sampling encryption mechanism; when the leakage level reaches the medium risk interval, the system sends warning information to the relevant post personnel and the ventilation control center through the wireless communication module, and triggers the wind flow direction optimization suggestion calculation; when the leakage level enters the high risk interval, the system automatically links the safety joint control system, issues an audible and visual alarm, and starts the emergency exhaust and personnel evacuation instructions.

[0060] In summary, the embodiments of the present application have at least the following technical effects: A sparse sensor network is deployed in a coal mine area to construct a coal mine cloud map; the sparse sensor network performs real-time data collection in the coal mine area and returns the data to the central control processor. If the real-time data indicates that there is no leakage but is in a metastable critical state, a mobile sensing component in the sparse sensor network is called to perform directional sensing management, wherein the leakage is determined based on the exceeding of the methane concentration of any sensing data. If the real-time data indicates that there is leakage, a leakage detection component is triggered, the data is synchronized to the coal mine cloud map, and cloud network screening and superposition state definition based on the leakage event are performed, a walking operator is introduced to collapse the superposition state based on the candidate leakage path, and leakage data containing the leakage source position and the target leakage path are determined and leakage warning management is performed. This solves the technical problem in the prior art that coal mine methane leakage monitoring relies on fixed sensing points and cannot dynamically track the leakage state, and achieves the technical effect of dynamic positioning of coal mine methane leakage.

[0061] Embodiment two, based on the same inventive concept as the coal mine methane leakage warning method based on a wireless sensor network in the foregoing embodiments, as shown in Figure 2 The present application provides a coal mine methane leakage warning system based on a wireless sensor network, wherein the system comprises: The cloud map construction module 11: deploy a sparse sensing network in the coal mine area, and construct a coal mine cloud map; the data acquisition module 12: the sparse sensing network performs real-time data acquisition of the coal mine area and returns to the central processor; the sensing management module 13: if the real-time data representation does not exist leakage, but in a metastable critical state, through the calling of the mobile sensing component in the sparse sensing network, directional sensing management is performed, wherein the leakage judgment is performed according to the exceeding of the methane concentration of any sensing data; the early warning module 14: if the real-time data representation exists leakage, trigger the leakage detection component, data synchronization to the coal mine cloud map and perform cloud network screening and superposition state definition based on the leakage event, through the introduction of the walking operator, based on the superposition state collapse of the candidate leakage path, determine the leakage data and perform leakage early warning management, wherein the leakage data includes the leakage source position and the target leakage path.

[0062] Further, the cloud map construction module 11 is used to perform the following method: Based on the sensing nodes of the sparse sensing network, the ventilation facility position, the potential leakage point and the monitoring blind area, the graph nodes are defined, wherein the sparse sensing network at least includes methane sensing type and environmental sensing type; the directed edge is defined according to the wind flow direction and the roadway connection relationship; the coal mine cloud map is constructed according to the graph nodes and the directed edge.

[0063] Further, the cloud map construction module 11 is used to perform the following method: Upload the coal mine condition variable, wherein the coal mine condition variable at least includes the working face opening, the ventilation route change; according to the coal mine condition variable, the coal mine cloud map is updated.

[0064] Further, the early warning module 14 is used to perform the following method: Based on the wind flow field of the ventilation network of the coal mine area, the real-time diffusion condition is determined; the real-time concentration signal of each methane sensing network is read, the leakage event network is screened in the coal mine cloud map, and each network node is initialized as a superposition state as a leakage event network; based on the leakage event network and the real-time diffusion condition, the leakage source positioning is performed.

[0065] Further, the early warning module 14 is used to perform the following method: Introduce the walking operator, and perform the leakage path exploration in the leakage event network under the constraint of the real-time diffusion condition, to determine the candidate leakage path set; perform the superposition state probability collapse processing on the candidate leakage path set to determine the leakage source position.

[0066] Further, the early warning module 14 is used to perform the following method: According to the real-time concentration signal, a source leakage mode is defined; if a first candidate leakage path does not match the source leakage mode, a path probability attenuation processing based on destructive interference is performed, wherein the first candidate leakage path is any one of the candidate leakage path set, and the attenuation scale is positively correlated with the degree of inconsistency; if the first candidate leakage path matches the source leakage mode to a higher degree, a path probability amplification processing based on constructive interference is performed, wherein the amplification scale is positively correlated with the degree of consistency.

[0067] Further, the early warning module 14 is used to perform the following method: According to the source leakage mode, a network node potential field vector based on the leakage event network is determined to generate a source leakage potential field; based on the source leakage potential field, a first comparison based on the potential field trend is performed on the first candidate leakage path, and a second comparison based on the potential field vector is performed, and the consistency analysis is jointly performed; wherein the path leakage source position of the target leakage path under extremely high probability collapse of the candidate leakage path set is taken as the leakage source position.

[0068] Further, the sensing management module 13 is used to perform the following method: A macroscopic behavior variable is defined, wherein the macroscopic behavior variable is defined according to the overall behavior description of the coal mine area, and at least includes concentration field information entropy, wind flow vortex intensity, and effective information flow of the sensing network; a dynamic sensing instruction is generated according to the macroscopic behavior variable, wherein the generation condition of the dynamic sensing instruction is the probability of the existence of a metastable critical state; a mobile sensing component in the sparse sensing network is called according to the dynamic sensing instruction to perform directional mobile sensing management; wherein the coal mine cloud map is updated synchronously with the change of the position of the mobile sensing component.

[0069] Further, the early warning module 14 is used to perform the following method: A unified three-dimensional coordinate system of the coal mine area is built; leakage identification information is generated based on the leakage data and written into the unified three-dimensional coordinate system as alarm data; and leakage early warning management based on the alarm data is performed.

[0070] The above is only a preferred embodiment of the present application, and does not limit the present application in any form. Although the preferred embodiment of the present application has been disclosed as above, it is not intended to limit the present application. Any person skilled in the art can make some changes or modifications to the above disclosed technical content to obtain equivalent embodiments with equivalent changes, without departing from the scope of the technical solution of the present application. Any modification, equivalent change and modification of the above embodiments based on the technical essence of the present application are still within the scope of the technical solution of the present application.

Claims

1. A method for early warning of methane leaks in coal mines based on wireless sensor networks, characterized in that, The method includes: Deploy sparse sensor networks in coal mining areas to construct coal mine cloud maps; The sparse sensor network collects real-time data from the coal mine area and transmits it back to the central control processor. If the real-time data representation does not show any leakage, but is in a metastable critical state, directional sensing management is performed by calling the mobile sensing component in the sparse sensor network. Leakage is determined by the methane concentration exceeding the limit of any sensor data. If real-time data indicates a leak, the leak detection component is triggered. The data is synchronized to the coal mine cloud map and cloud map network screening and superposition state definition are performed based on the leak event. By introducing a walking operator, the leak data is determined and leak early warning management is performed based on the superposition state collapse of candidate leak paths. The leak data includes the location of the leak source and the target leak path.

2. The method for early warning of methane leakage in coal mines based on wireless sensor networks as described in claim 1, characterized in that, Deploying sparse sensor networks in coal mining areas to construct coal mine cloud maps includes: Graph nodes are defined based on sensor nodes, ventilation facility locations, potential leak points, and monitoring blind spots in a sparse sensor network. The sparse sensor network includes at least methane sensing and environmental sensing types. Directed edges are defined based on airflow direction and the connection relationship between alleyways; The coal mine cloud graph is constructed based on the graph nodes and the directed edges.

3. The method for early warning of methane leakage in coal mines based on wireless sensor networks as described in claim 2, characterized in that, After constructing the coal mine cloud map, the following is included: Upload coal mine condition variables, wherein the coal mine condition variables include at least the opening of the working face and the change of ventilation route; The coal mine cloud map is updated based on the coal mine condition variables.

4. The method for early warning of methane leakage in coal mines based on wireless sensor networks as described in claim 1, characterized in that, Identify the leaked data, including: Based on the airflow field of the ventilation network in the coal mine area, the real-time diffusion conditions are determined. Read the real-time concentration signals of each methane sensor network, filter the leakage event network in the coal mine cloud map, and initialize a superposition state for each network node as the leakage event network; Based on the leakage event network and the real-time diffusion conditions, the leakage source is located.

5. The method for early warning of methane leakage in coal mines based on wireless sensor networks as described in claim 4, characterized in that, Based on the leakage event network and the real-time diffusion conditions, the leakage source is located, including: A walking operator is introduced, and with the real-time diffusion condition as a constraint, leakage paths are explored in the leakage event network to determine a set of candidate leakage paths. The candidate leakage path set is subjected to superposition state probability collapse processing to determine the location of the leakage source.

6. The method for early warning of methane leakage in coal mines based on wireless sensor networks as described in claim 5, characterized in that, Before performing superposition-state probability collapse processing on the candidate leakage path set, the following steps are included: Based on the real-time concentration signal, define the source leakage mode; If the first candidate leakage path does not match the source leakage mode, path probability attenuation processing based on destructive interference is performed, wherein the first candidate leakage path is any one of the candidate leakage paths in the set, and the attenuation scale is positively correlated with the degree of inconsistency; If the first candidate leakage path matches the source leakage mode to a high degree, path probability amplification processing based on constructive interference is performed, wherein the amplification scale is positively correlated with the degree of matching.

7. The method for early warning of methane leakage in coal mines based on wireless sensor networks as described in claim 6, characterized in that, Perform superposition-state probability collapse processing on the candidate leakage path set to determine the leakage source location, including: Based on the source leakage pattern, determine the network node potential field vector based on the leakage event network, and generate the source leakage potential field; Based on the source leakage potential field, a first comparison based on the potential field trend and a second comparison based on the potential field vector are performed on the first candidate leakage path, and a joint consistency analysis is conducted. The location of the leak source is determined by the path leakage source of the target leak path under a very high probability collapse of the candidate leak path set.

8. The method for early warning of methane leakage in coal mines based on wireless sensor networks as described in claim 1, characterized in that, Directional sensing management is achieved by invoking mobile sensing components within a sparse sensor network, including: Define macroscopic behavioral variables, wherein the macroscopic behavioral variables are defined according to the overall behavioral description of the coal mining area, and include at least concentration field information entropy, wind vortex intensity, and effective information flow of sensor network; Based on the macroscopic behavioral variables, dynamic sensing commands are generated, wherein the generation condition of the dynamic sensing commands is the probability of the existence of a metastable critical state. According to the dynamic sensing command, the mobile sensing component in the sparse sensor network is invoked to perform directional mobile sensing management. The coal mine cloud map is updated synchronously as the position of the mobile sensing component changes.

9. The method for early warning of methane leakage in coal mines based on wireless sensor networks as described in claim 1, characterized in that, Leakage early warning management includes: Establish a unified three-dimensional coordinate system for the coal mining area; Based on the leaked data, leak identification information is generated and written into the unified three-dimensional coordinate system as alarm data; Perform leakage early warning management based on the alarm data.

10. A coal mine methane leak early warning system based on a wireless sensor network, characterized in that, For implementing the coal mine methane leak early warning method based on a wireless sensor network according to any one of claims 1-9, the system comprises: Cloud map construction module: Deploy sparse sensor networks in coal mining areas to construct coal mine cloud maps; Data acquisition module: The sparse sensor network performs real-time data acquisition of the coal mine area and transmits it back to the central control processor; Sensor Management Module: If the real-time data characterization does not show any leakage, but is in a metastable critical state, directional sensor management is performed by calling the mobile sensor component in the sparse sensor network. Leakage is determined by the methane concentration exceeding the limit of any sensor data. Early warning module: If real-time data indicates a leak, the leak detection component is triggered, the data is synchronized to the coal mine cloud map and cloud map network screening and superposition state definition are performed based on the leak event. By introducing the walking operator, the leak data is determined and leak early warning management is performed based on the superposition state collapse of the candidate leak path. The leak data includes the location of the leak source and the target leak path.

Images