A data acquisition method for explosion-proof equipment in an explosion-hazardous place

Abstract

The present application relates to the technical field of data acquisition for explosion-proof equipment in explosion-hazardous places, and discloses a data acquisition method for explosion-proof equipment in explosion-hazardous places. A three-dimensional Cartesian coordinate system is established to obtain the spatial positions of nodes, and a plurality of sub-regions are divided according to the spatial structure to complete node attribution mapping. The failure probability of nodes and communication links is calculated through historical data, and a network topology is constructed in combination with communication reachability to determine the optimal reliable path based on node and link cost functions. Subsequently, regional redundancy planning is carried out, and the main nodes and backup nodes are selected according to the path cost; the sampling period is divided into time slots and the main node sampling time is reasonably allocated; the main node failure detection and backup node replacement are realized during system operation; after the period ends, the main node uploads the collected data to the sink node and completes the integrity check.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data acquisition for explosion-proof equipment in explosion-hazardous places, and particularly to a data acquisition method for explosion-proof equipment in explosion-hazardous places. BACKGROUND

[0002] With the acceleration of industrialization process and the increasing requirement of safety supervision, the environmental monitoring of explosion-hazardous places becomes particularly important. In order to timely find dangerous signals such as combustible gas leakage, temperature anomaly, pressure fluctuation, a large number of wireless sensor nodes are usually deployed in the dangerous area, and the measurement data of each sensor is transmitted in real time to the central monitoring system through a wireless link. However, the prior art has many deficiencies in such a scene, and it is difficult to meet the continuous, reliable, accurate and orderly data acquisition requirements.

[0003] Firstly, the prior art lacks accurate and unified coordinate reference in spatial positioning. Most systems only rely on field installation records or approximate estimates of node positions, making it difficult to accurately describe the actual position of sensor nodes in three-dimensional space. This leads to uncertainty in the distance between different nodes, affecting the construction of data transmission paths and network topology, and making it impossible to reasonably partition and manage dangerous areas based on node positions. Secondly, existing wireless data collection schemes mainly rely on a single communication link or a pre-set path, without fully considering the failure probability of nodes and links. In explosive hazardous environments, sensor nodes are often exposed to harsh conditions such as high temperature, humidity, and corrosive gases, increasing the risk of hardware failure. Meanwhile, wireless communication is easily affected by obstructions and interference, but the lack of link reliability is often overlooked. The lack of quantitative calculation of single-point failure probability and inter-node link failure rate makes the network prone to data loss and path interruption when node failures or unstable links occur, making it difficult to ensure the timely transmission of critical alarm information. The existing technology lacks sufficient network redundancy planning and lacks differentiated redundancy strategies for different sub-regions. Traditional methods often simply increase the number of devices in areas with high device density or where damage is easy, but lack quantitative evaluation based on node and link reliability, and lack planning of the primary and backup relationship within the region. As a result, it is difficult to quickly replace nodes in different sub-regions after a node failure, and the system's recovery ability is poor once the primary node is lost, posing a significant safety risk. In terms of time slot allocation and sampling scheduling, traditional systems mostly use fixed-cycle synchronous polling or random access strategies. Fixed polling is easy to implement, but it is prone to conflicts and delays; while random access cannot guarantee real-time and stability. In explosive hazardous places, the timeliness requirement for sampling data is extremely high, and any conflict or transmission delay may miss the alarm opportunity, but existing solutions lack fine scheduling design that divides the sampling period into multiple time slots and avoids conflicts for different primary nodes. The existing solutions also lack simple fault detection and switching mechanisms. The common approach is to conduct periodic diagnosis or manual inspection, and replace the node only when it is found to be offline or data is lost, and lacks automatic judgment logic based on the number of consecutive packet losses, often missing the best switching opportunity. Manual inspection not only consumes time and effort, but also may miss the short window when the node fails, resulting in important data not being reported in a timely manner. Data integrity verification is often simply a matter of determining whether a data packet has been received, lacking comprehensive evaluation of data integrity at the regional and network levels. Clear regional integrity indicators and network integrity indicators are not established to accurately identify which sub-regions have data missing and alarm in a timely manner during program execution. Traditional monitoring systems often only focus on whether each node is online, but do not focus on whether the entire network covers all primary node data within a certain period, lacking a comprehensive understanding of data quality and availability.

[0004] Therefore, the present application aims to provide a data acquisition method for explosion-proof equipment in explosion-hazardous places, and aims to build a highly reliable, recoverable, time-effective and integrity-detectable distributed sensor network system. Precise positioning and regional division are established based on a three-dimensional coordinate system, physical deployment management of each node is realized, robust topology structure and path planning are built through historical data-driven fault modeling, node-level fault tolerance is realized by using redundancy and backup mechanism, and data ordered uploading and comprehensive acquisition are ensured through strict time slot management and integrity detection. SUMMARY

[0005] The present application provides a data acquisition method for explosion-proof equipment in explosion-hazardous places, which solves the problems mentioned in the background art.

[0006] The present application provides the following technical solution: a data acquisition method for explosion-proof equipment in explosion-hazardous places, comprising:

[0007] A three-dimensional Cartesian coordinate system of the scene is established, and the three-dimensional coordinates of each sensor node are obtained;

[0008] The scene is divided into multiple sub-regions according to the spatial structure, and each node is attributed to the corresponding region;

[0009] The failure probability of the sensor node and the failure probability of the communication link are calculated based on historical monitoring data;

[0010] The network topology structure is built based on the node communicability judgment, and the optimal reliable path of each node to the sink node is determined according to the cost function of the node and the link;

[0011] Redundancy planning is performed for each sub-region, and the master node and the backup node are selected based on the path cost;

[0012] A sampling period is divided into multiple time slots, and each master node is allocated a specific sampling time slot;

[0013] Master node failure detection is performed during operation, and the backup node takes over the sampling task when a failure occurs;

[0014] The collected data is sent by the master node to the sink node, and data integrity verification is performed after the period ends.

[0015] Optionally, the three-dimensional Cartesian coordinate system of the scene is established, and the three-dimensional coordinates of each sensor node are obtained, specifically comprising:

[0016] The explosion scene is a rectangular space, the bottom surface is a rectangle containing two long sides and two short sides, and the length of the long side is greater than the length of the short side;

[0017] The center point of the bottom surface is obtained as the origin of the coordinate system, denoted as ;

[0018] Through the original point, a ray parallel to the long side of the bottom surface is made, and the ray intersects the two short sides of the bottom surface respectively. The ray parallel to the long side of the bottom surface is taken as the x-axis of the coordinate system, and the direction of the ray is the positive direction of the x-axis. Through the original point, a ray parallel to the short side of the bottom surface is made, and the ray intersects the two long sides of the bottom surface respectively. The ray parallel to the short side of the bottom surface is taken as the y-axis of the coordinate system, and the direction of the ray is the positive direction of the y-axis.

[0019] Through the original point, a ray perpendicular to the bottom surface is made, and the direction of the ray is vertically upward. The ray intersects the upper and lower bottom surfaces of the rectangular space respectively. The ray perpendicular to the bottom surface is taken as the z-axis of the coordinate system, and the direction of the ray is the positive direction of the z-axis.

[0020] In the coordinate system, a plurality of sensor nodes are deployed, numbered as 1, 2, …, n, and the three-dimensional coordinates of each node are represented as (x i, y i, z i), i = 1, 2, …, n.

[0021] In the coordinate system, a plurality of sensor nodes are deployed, numbered as 1, 2, …, n, and the three-dimensional coordinates of each node are represented as (x i, y i, z i), i = 1, 2, …, n. is the coordinate value of the node on the x-axis; is the coordinate value of the node on the y-axis; is the coordinate value of the node on the z-axis.

[0022] The sensors are all of the same model.

[0023] A sink node is set, numbered as n + 1.

[0024] The three-dimensional coordinates of the sink node are represented as (x s, y s, z s). is the coordinate value of the sink node on the x-axis; is the coordinate value of the sink node on the y-axis; is the coordinate value of the sink node on the z-axis.

[0025] Optionally, the space is divided into a plurality of sub-regions according to the space structure, and each node is attributed to a corresponding region, specifically including:

[0026] In the coordinate system, the explosion site is divided into​​​​​​​​​​​​​​​​​​​​ a cuboid sub-region numbered as ;

[0027] The first region is represented by the following six parameters, specifically:

[0028] ; wherein, and are the minimum coordinate and the maximum coordinate of the first region on the axis respectively; and are the minimum coordinate and the maximum coordinate of the first region on the axis respectively; and are the minimum coordinate and the maximum coordinate of the first region on the axis respectively;

[0029] The node attribution function is set as:

[0030] ;

[0031] When the node coordinate is on the boundary of multiple regions, the region with the smallest number is attributed;

[0032] The node set in the first region is set as:

[0033] , ; wherein, is the number of nodes in the first region; is the number of elements in the set .

[0034] Optionally, the calculation of the failure probability of the sensor node and the failure probability of the communication link based on the historical monitoring data specifically includes:

[0035] The first sensor is continuously monitored, and the total monitoring sampling period is set as , and the number of missing data of the first node in the monitoring period is recorded as ;

[0036] For each sensor node , the single-point failure rate is set as , and satisfies ;

[0037] The sink node is denoted as ;

[0038] For each pair of nodes , the three-dimensional Euclidean distance between the two points is calculated:

[0039] , ;

[0040] The maximum distance of communication of the sensor node is denoted as , and is taken as the effective distance threshold of communication between nodes;

[0041] The nodes may communicate only when and , otherwise there is no communication link;

[0042] If , the failure probability of the communication link between the nodes and in one transmission process is set as ;

[0043] The communication reachability indication is set as:

[0044] ; wherein, indicates that the nodes and may communicate; indicates that the nodes and cannot communicate; .

[0045] Optionally, the network topology is constructed on the basis of the node communicability judgment, and the optimal reliable path of each node to the sink node is determined according to the cost function of the nodes and the links, and specifically includes:

[0046] The set of all nodes is set as ; wherein, the number 0 represents the sink node, and the numbers 1 to represent each sensor node;

[0047] The edge set is set as ; wherein, indicates that the nodes and have an available link in one communication cycle;

[0048] The undirected graph is constructed as ;

[0049] For each node , set the node cost of ; wherein, ;

[0050] For each edge , set the edge cost of , ;

[0051] For any node , let a simple path from node to the sink node be:

[0052] ; wherein, denotes the th node on the path; and all are not repeated; is the number of intermediate nodes on the path from node to the sink node;

[0053] Set the total reliability of the path as:

[0054] ;

[0055] Set the total cost of the path as:

[0056] ; wherein, is the edge cost of the th edge; is the node cost of the th node;

[0057] Set the optimal path search algorithm as:

[0058] S100, initialization:

[0059] S101, set the array as the current minimum path cost from node to the sink node:

[0060] , ;

[0061] S102, set the predecessor array initialized as undefined;

[0062] S103, set the set of nodes with fixed optimal cost ;

[0063] S200, iteration process:

[0064] When is not empty, repeat the following operations:

[0065] S201, Order , will node Merge into an existing fixed set ;like If the condition is met, the loop exits and the algorithm terminates; otherwise, it continues.

[0066] S202, for each and There are connected edges and none have been added. nodes Perform a relaxation update:

[0067] Calculation passed Total cost as a relay to the aggregation node ;

[0068] like Then update , ;

[0069] Repeat step S200 until all nodes are added. Or the remaining nodes are unreachable;

[0070] S300, Path Backtracking and Result Output:

[0071] For each number ,like Then through array from By backtracking step by step to the convergence node, the optimal path is obtained. ;at the same time, This is the total cost of the optimal path;

[0072] like , then it represents a node There is no reachable path to the aggregation node.

[0073] Optionally, the redundancy planning for each sub-region and the selection of primary and backup nodes based on path cost specifically includes:

[0074] For the The region is configured with the following number of master nodes: ;in, It is a rounding function;

[0075] In the In the region, for all nodes According to path cost Sort the nodes in ascending order from smallest to largest, and denote the sorted node index sequence as follows: ,and ;

[0076] Set the first The set of master nodes for the region is:

[0077] ;

[0078] Set the first The set of spare nodes for the region is:

[0079] ;

[0080] like The number of backup nodes ;

[0081] For sets All nodes are assigned a spare activation flag. Initialize to 0; where, if If it is, then it is considered the master node. Keep it at 0.

[0082] Optionally, dividing a sampling period into multiple time slots and assigning specific sampling time slots to each master node specifically includes:

[0083] Set the sampling period to ;

[0084] And Divided into equal parts There are 1 time slot, and the length of each time slot is:

[0085] ;

[0086] For the first The first in the region One master node Candidate time slot numbers are defined sequentially as follows:

[0087] ;in, This is the floor function;

[0088] Set the global time slot occupancy marker array to Initialize all values ​​to 0;

[0089] By region number Perform the following operations in sequence:

[0090] S400, for the region Each master node in Set candidate time slot number ;

[0091] S500, if Then let the allocation function and will ;

[0092] S600、If , from start sequentially increasing, if increased to , let , check until the time slot value is 0, and then assign the time slot to and mark ;

[0093] S700, until the first master node time slot allocation is completed, continue to the next master node;

[0094] Finally, each master node has a unique sampling time slot in each cycle , which is , corresponding to the actual sampling time , at which time the node collects a measurement value and sends it to the sink node; wherein is the cycle index.

[0095] Optionally, the master node fault detection is performed during operation, and the standby node replaces the sampling task when a fault occurs, specifically including:

[0096] Set the receiving flag of the node in the first cycle as:

[0097] ;

[0098] wherein is the sequence number of the node in its regional master node set; is the master node set of the first sub-region; is the receiving flag of whether the sink node has received the data of the master node in the first sampling cycle;

[0099] Set the fault counter of the first node at the end of the first cycle as:

[0100] , ; and

[0101] Set the continuous fault determination threshold as ;

[0102] When , the node is determined to be faulty;

[0103] If there is a master node at the end of the cycle satisfying , the following replacement operation is performed:

[0104] S810, remove from the master node set, specifically: ;

[0105] S820, in the standby node set , find the first node that is not activated in order from the smallest index to the largest , that is, satisfying ;

[0106] S830, add the node to the master node set, specifically:

[0107] , ;

[0108] S840, assign a time slot to the newly added node :

[0109] S841, let the candidate time slot number be ;

[0110] S842, if ,

[0111] , let , and update ;

[0112] wherein is the standby node in decrement 1 in ranking;

[0113] S843, otherwise, sequentially increase to the next free time slot until a free time slot is found and marked;

[0114] S850, initialize the failure counter of the new replacement node : ;

[0115] If there is no available standby node in , the area only retains the remaining master node to continue the work.

[0116] Optionally, the collected data is sent by the master node to the sink node, and data integrity check is performed after the cycle ends, specifically including:

[0117] For the first each master node in a period at time sending measurement values The situation of receiving at the sink node is set as:

[0118] ;

[0119] constructing the first network sampling data matrix in the period :

[0120] ;

[0121] ; wherein, is a sampling result record three tuple of the master node in the first sampling period;

[0122] The data integrity index of the first region in the first period is set as:

[0123] ;

[0124] The data integrity index of the first period of the whole network is set as:

[0125] ; when and only when all , , otherwise the value of is 0;

[0126] The sink node synchronously stores the following information after the end of each period and can upload it to the monitoring center on demand: , .

[0127] The present application has the following beneficial effects:

[0128] 1. A precise three-dimensional Cartesian coordinate system is introduced in an explosion hazard place to uniformly manage the positions of sensor nodes and sink nodes. Unlike the traditional two-dimensional deployment method, the three-dimensional modeling can more realistically reflect the actual layout environment of the equipment, and is particularly suitable for industrial places with multi-layer and multi-height structures. The bottom center is set as the origin to ensure the spatial geometric balance; the coordinate axis direction is accurately defined, so that the distance between nodes, region division and subsequent path calculation have mathematical rigor. Through clear numbering and coordinate recording, the spatial index efficiency is significantly improved, the possibility of node deployment confusion is reduced, and a solid foundation is laid for subsequent partitioning, topology and communication design.

[0129] 2、Proposed a rectangular cuboid unit-based regional partition model, combined with the coordinate axis boundary to automatically attribute the node. Provide mathematical definition of attribution function, make the node attribution completely rely on spatial location, avoid subjective division or artificial configuration; when dealing with boundary coincident nodes, adopt the rule of minimum number priority, avoid multi-attribution conflict. This partition method supports fast positioning of regional node set, and is convenient for independent scheduling and redundant configuration within the region. Most existing schemes use static partition based on logical groups, lack of spatial boundary constraints, leading to inconsistency between partition and actual physical topology; this scheme makes the partition have explainability and traceability through spatial geometric model, significantly improves the efficiency of sub-region management and the controllability of node distribution.

[0130] 3、The scheme uses actual historical data sampling rate statistics to construct a node failure probability model, and combines three-dimensional space distance calculation to calculate the communication link failure rate, realizing a communication judgment mechanism with precision and effectiveness. Node failure is not a preset or subjective threshold, but a dynamic learning from historical data; the failure of communication link uses a probability model that is a function of distance, reflecting the physical reality of wireless environment, supporting dynamic adjustment of network.

[0131] 4、Proposed a path cost function based on node and link reliability, and calculated the optimal path to the sink node using an improved shortest path algorithm. The reliability (probability product) is converted into cost and (logarithm sum) to adapt to the shortest path model, so that various uncertain factors are integrated into a unified optimization framework; the predecessor array and path backtracking mechanism are introduced based on the graph model to ensure clear and traceable path output. Traditional path algorithms mostly consider distance or hop count, ignoring actual failure probability; while this method integrates reliability calculation, making the path selection not only short but also stable, especially suitable for continuous data transmission in extreme environments, reducing the overall transmission failure risk.

[0132] 5、By setting regional redundancy factor and ranking nodes by total path cost, a primary and backup node set is constructed. The primary and backup nodes are selected based on global path performance rather than physical proximity, improving the continuity of system performance after the backup node takes over; the backup node activation flag mechanism realizes dynamic management, avoiding unnecessary switching and resource waste. Traditional network redundancy design mostly uses fixed backup nodes and fixed regional allocation, which cannot dynamically respond to path or node failure; while this scheme combines path stability and actual node density in the region to design a more intelligent and efficient primary and backup configuration method.

[0133] 6、Adopting isochronous slot division sampling period, combined with candidate time slot list and global conflict flag table to realize primary node time slot allocation. The time slot selection algorithm ensures that there is no interference between primary nodes within the region and that the distribution across regions is as uniform as possible, improving the overall coordination of transmission timing; the conflict detection and dynamic allocation mechanism is optimal and flexible, adapting to the reallocation needs after the number of nodes changes.

[0134] 7. The stability of the node is determined by periodically receiving a flag and a failure counter, and the determination threshold is set to control the switching condition. The tolerance mechanism (two consecutive failure determinations as a failure) is introduced to improve the anti-jitter ability of the system; the backup node sequential search, dynamic replacement and time slot reallocation mechanism constitute a closed loop process to ensure the self-healing ability of the system. The traditional system is mostly manually intervened or restarted to take effect, which has high delay and is not intelligent; the present scheme has high autonomy and fast response characteristics, and is especially suitable for high-risk places to ensure data continuity and security in emergency situations.

[0135] 8. After the end of the collection period, the system constructs a full network data matrix and calculates the regional and full network integrity indicators, realizing a multi-dimensional data integrity evaluation mechanism. Not only single-point anomaly detection is supported, but also regional diagnosis is supported to improve the alarm accuracy; the integrity result is used as the basis for uploading to the monitoring center to support intelligent decision-making and maintenance scheduling. The conventional system often only pays attention to whether it is "received", but ignores regional coverage and structural completeness; the present method expands the definition of integrity to the structural level, making the data reliability evaluation more systematic and scientific, and providing a high-quality data source for the back-end processing. BRIEF DESCRIPTION OF DRAWINGS

[0136] Figure 1 The present application is a flowchart.

[0137] Figure 2 The present application is a three-dimensional Cartesian coordinate system diagram.

[0138] In the figure: 1-rectangular space, 2-bottom surface, 3-long side, 4-short side, 5-origin, 6- axis, 7- axis, 8- axis. DETAILED DESCRIPTION

[0139] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of the present application.

[0140] Embodiment, refer to Figure 1 A data collection method for explosion-proof equipment in an explosion-hazardous place, comprising:

[0141] establishing a three-dimensional Cartesian coordinate system of the site and obtaining three-dimensional coordinates of each sensor node;

[0142] dividing the site into a plurality of sub-regions according to the spatial structure, and attributing each node to the corresponding region;

[0143] calculating the failure probability of the sensor nodes and the failure probability of the communication links based on historical monitoring data;

[0144] constructing a network topology based on the node communicability judgment, and determining the optimal reliable path of each node to the sink node according to the cost function of the nodes and links;

[0145] planning the redundancy for each sub-region, and selecting the primary node and the backup node based on the path cost;

[0146] dividing one sampling period into multiple time slots, and allocating specific sampling time slots to each primary node;

[0147] detecting the primary node failure during the operation, and replacing the primary node with the backup node to perform the sampling task when a failure occurs;

[0148] sending the collected data from the primary node to the sink node, and performing data integrity verification after the end of the period.

[0149] A complete data acquisition process is provided, which is suitable for the monitoring system of explosion-proof equipment in an explosion hazard site. By establishing a three-dimensional coordinate system on site, the actual space is accurately mapped into a digital model, laying a foundation for node layout and path planning; sub-regions are divided and nodes are assigned to belong to, realizing regional management, facilitating local optimization and fault tolerance configuration; historical failure probability and communication link reliability analysis are introduced, providing dynamic adjustable reference factors for topology construction; on this basis, combined with the primary and backup node mechanism and the optimal path search algorithm, the robustness and high reliability of the data acquisition process are ensured; in combination with the time slot allocation mechanism and the failure replacement mechanism, the system has stable data acquisition rhythm and self-recovery ability; finally, the effectiveness and coverage of the data are verified through the integrity verification mechanism. Through a series of steps, the problems of unordered node deployment, unstable communication and high risk of data loss in an explosion hazard environment are effectively solved, and the safe, stable and efficient operation of the system is realized.

[0150] Referring to Figure 2 , the three-dimensional Cartesian coordinate system of the site is established, and the three-dimensional coordinates of each sensor node are obtained, specifically including:

[0151] The explosion site is a rectangular space with a bottom surface that is a rectangle containing two long sides and two short sides, and the length of the long side is greater than the length of the short side;

[0152] The center point of the bottom surface is obtained as the origin of the coordinate system, denoted as ; a unified zero reference is constructed for the entire site, so that all subsequent space coordinate calculations have a clear starting point;

[0153] A ray is made through the origin and parallel to the long side of the bottom surface, and the ray intersects the two short sides of the bottom surface respectively. The ray parallel to the long side of the bottom surface is taken as the axis of the coordinate system, and the direction of the ray is the positive direction of the axis.

[0154] A ray is made through the origin and parallel to the short side of the bottom surface, and the ray intersects the two long sides of the bottom surface respectively. The ray parallel to the short side of the bottom surface is taken as the axis of the coordinate system, and the direction of the ray is the positive direction of the axis.

[0155] A ray is made through the origin and perpendicular to the bottom surface, and the direction of the ray is vertically upward. The ray intersects the upper and lower bottom surfaces of the rectangular space respectively. The ray perpendicular to the bottom surface is taken as the axis of the coordinate system, and the direction of the ray is the positive direction of the axis.

[0156] The direction of the coordinate axis is defined to ensure that all node coordinate measurements and subsequent geometric operations are performed under the same reference.

[0157] In the coordinate system, a plurality of sensor nodes are deployed , numbered . The sensor nodes are numbered for subsequent indexing and management. The three-dimensional coordinates of each node are represented as ; wherein is the coordinate value of the node on the axis; is the coordinate value of the node on the axis; is the coordinate value of the node on the axis. The accurate positions of the sensor nodes in the three-dimensional space are recorded, providing a basis for subsequent distance and area division.

[0158] The sensors are all of the same model.

[0159] A sink node is set, numbered .

[0160] The three-dimensional coordinates of the sink node are denoted as ; wherein is the coordinate value of the sink node on the axis; is the coordinate value of the sink node on the axis; is the coordinate value of the sink node on the The coordinate value on the axis; set the number and coordinate for the data aggregation endpoint, which is convenient for network topology construction as a target node.

[0161] By establishing a three-dimensional Cartesian coordinate system and clearly defining the three-dimensional position of each node, the precise modeling of the entire collection network space structure is realized. Its definition method with the bottom center as the origin, the long / short direction as the X / Y axis, and the vertical direction as the Z axis can realize a unified spatial coordinate system for rectangular explosion sites. Node numbering and coordinate recording further enhance the standardization and controllability of space management. Through this step, the problems of node position ambiguity, insufficient positioning accuracy, and difficult spatial scheduling in traditional wireless collection systems are effectively solved; at the same time, it provides a strict geometric basis for subsequent regional division, link distance judgment, and path planning, improving the engineering realizability and expansibility of system modeling.

[0162] The field is divided into multiple sub-regions according to the spatial structure, and each node is attributed to the corresponding region, specifically including:

[0163] In the coordinate system, the explosion site is divided into non-overlapping cuboid sub-regions with parallel boundaries and coordinate axes, numbered ; the large site is divided into several sub-regions for easy management and monitoring, facilitating zoned arrangement and zoned monitoring;

[0164] The first region is represented by the following six parameters, specifically:

[0165] ; wherein and are the minimum and maximum coordinates of the first region on the axis; and are the minimum and maximum coordinates of the first region on the axis; and are the minimum and maximum coordinates of the first region on the axis; provide accurate spatial boundary description for judging whether a node falls into this region;

[0166] The node attribution function is set as:

[0167] ; each sensor node is automatically mapped to its sub-region, facilitating subsequent zoned management;

[0168] When the node coordinates are on the multiple region overlapping boundary, the region with the smallest number The home is; The specific implementation node to area belonging judgment rule, guarantee each node unique belonging one sub area;

[0169] Set the first The node set in the area is:

[0170] , ; wherein, The number of nodes in the first area; The number of elements of the set ; Facilitate subsequent fast get all the node index of the first area for processing; Statistics and record the number of sensor nodes in each area, provide basic data for redundancy planning.

[0171] Through the six parameters definition of three-dimensional space, the accurate partition of explosion site is realized, and the sub area structure based on space logic is constructed. Each node is automatically attributed to a certain area according to its coordinate value, and the smallest number priority strategy is adopted when the boundary coincides, so as to avoid attribution conflict. The area division mechanism not only improves the hierarchy and modularity of node management, but also provides data and structure basis for subsequent master and standby node redundancy configuration, time slot scheduling and fault isolation. Through this step, the problems such as data scheduling complexity, operation and maintenance coverage confusion caused by "global flat arrangement" in traditional system are solved, the modularization manageability of the system is effectively improved, and the maintenance cost and operation and maintenance complexity are reduced.

[0172] The fault probability of the sensor node and the fault probability of the communication link are calculated based on the historical monitoring data, and specifically include:

[0173] The first sensor is continuously monitored, and the total monitoring sampling period is , the number of missing data of the first node in the monitoring period is recorded as ;

[0174] For each sensor node , the single point failure rate is , and satisfies ;

[0175] The sink node is denoted as ;

[0176] The possibility of each node itself appearing fault and losing data is quantified, which provides basic parameters for subsequent path reliability calculation;

[0177] For each pair of nodes , the three-dimensional Euclidean distance between the two points is calculated:

[0178] , ; for subsequent determination of whether two nodes are within the communication range and calculation of link failure probability;

[0179] The maximum distance of sensor node communication is denoted as , and is set as the effective distance threshold of inter-node communication; the maximum distance of inter-node communication is limited;

[0180] Only when , the node may communicate with the node , otherwise there is no communication link;

[0181] If , the failure probability of the communication link between the node and the node in one transmission process is set as ; a quadratic function relationship is expressed that the closer the distance between the node pair, the more reliable the link, and the farther the distance, the greater the failure rate of the link, which is used for subsequent calculation of path reliability;

[0182] The communication reachability indicator is set as:

[0183] ; wherein, indicates that the node may communicate with the node ; indicates that the node cannot communicate with the node ; ; for quickly determining whether there is a usable communication link between the node pair, and constructing the edge condition of the network topology.

[0184] By monitoring the missing situation of the sampling historical data, the single-point failure rate of each node is counted, and the failure probability of the communication link is set in combination with the three-dimensional Euclidean distance between the nodes, thereby constructing a complete failure modeling mechanism. The communication reachability indicator matrix can be directly used for the construction of the network topology. The method fuses data driving and geometric distance, and no longer relies on artificial experience to set the network connectivity and stability, greatly improving the adaptive ability of the model. The mechanism solves the problems of single failure detection and link modeling and poor accuracy in the existing system, makes the construction of the network topology more reasonable and robust, and effectively supports the subsequent path optimization and fault-tolerant configuration.

[0185] The network topology is constructed on the basis of the node communicability determination, and the optimal reliable path of each node to the sink node is determined according to the cost function of the node and the link, specifically including:

[0186] Set the set of all nodes as ; where, number 0 represents the sink node, and numbers 1 to N represent the sensor nodes; ; represents each sensor node; represents all nodes in the network, including the sink node 0 and the sensor nodes 1 to N;

[0187] Set the set of edges as ; where, represents that node has an available link with node in one communication cycle; record all communicable node pairs for subsequent path search;

[0188] Build the undirected graph as ;

[0189] For each node , set the node cost as ; where, ; convert the single-point failure probability into a logarithmic cost, so that the total path cost can be accumulated;

[0190] For each edge , set the edge cost as , ; convert the link failure probability into a logarithmic cost, which is convenient for accumulation with the same dimension as the node cost;

[0191] For any node , set a simple path from node to the sink node as:

[0192] ; where, represents the th node on the path; and all are not repeated; is the number of intermediate nodes on the path from node to the sink node; define the node order of the path to ensure that all indices are not repeated;

[0193] Set the total reliability of the path as:

[0194] ; combine the node and link reliabilities to quantify the probability that the entire path does not fail in one sampling transmission process;

[0195] Set the total cost of the path as:

[0196] ; where, is the edge cost of the th edge; is the The node cost of each node; taking the negative logarithm of the overall reliability transforms the multiplicative relationship into an summable cost, which facilitates the calculation of the shortest path;

[0197] Set the optimal path search algorithm as follows:

[0198] S100, Initialization:

[0199] S101, Set array For the node The current minimum path cost to the convergence node:

[0200] , During initialization, only the cost of reaching a node from itself is 0; the rest are unknown or unreachable.

[0201] S102, Set the predecessor array Initialize to undefined; record nodes in the optimal path. The next node index is used to facilitate the final backtracking path;

[0202] S103. Assume that the set of optimal cost nodes has been fixed. Mark nodes that have determined the optimal total cost and completed the relaxation operation to avoid duplicate processing;

[0203] S200, Iterative Process:

[0204] exist If not empty, repeat the following operations:

[0205] S201, Order , will node Merge into an existing fixed set ;like If the condition is met, the loop exits and the algorithm terminates; otherwise, it continues.

[0206] Greedily select the most reliable node from the unprocessed nodes and add it to the "determined" set;

[0207] S202, for each and There are connected edges and none have been added. nodes Perform a relaxation update:

[0208] Calculation passed Total cost as a relay to the aggregation node ;Evaluate if the path contains nodes Then cross the border arrive The total cost is compared with the currently known total cost;

[0209] like Then update , If the new total cost is smaller, replace the old value and record the predecessor node as... This is to prepare for subsequent path backtracking;

[0210] Repeat step S200 until all nodes are added. Or the remaining nodes are unreachable;

[0211] S300, Path Backtracking and Result Output:

[0212] For each number ,like Then through array from By backtracking step by step to the convergence node, the optimal path is obtained. ;at the same time, This is the total cost of the optimal path;

[0213] like , then it represents a node There is no reachable path to the aggregation node;

[0214] By backtracking the optimal path step by step using the predecessor array, a complete path sequence can be obtained.

[0215] By constructing an undirected graph structure, introducing node cost (logarithm of failure probability) and edge cost (logarithm of link failure), and combining it with a shortest path search algorithm after reliability transformation, the optimal path calculation from node to sink node is achieved. This algorithm selects the path with the lowest cost (i.e., lowest failure probability) while ensuring communication connectivity, and provides a predecessor index for path backtracking and result application. This technology effectively solves the problems of unstable paths and high transmission failure rates in explosive environments. Compared to traditional methods that only consider the shortest hop count or physical distance, this method significantly improves path stability and overall system reliability, and is a key foundation for realizing a highly robust data acquisition system.

[0216] The redundancy planning for each sub-region and the selection of primary and backup nodes based on path cost specifically includes:

[0217] For the The region is configured with the following number of master nodes: ;in, It is a rounding function; while ensuring redundancy, it covers the area with the fewest number of master nodes, improving efficiency and reliability.

[0218] In the In the region, for all nodes According to path cost Sort the nodes in ascending order from smallest to largest, and denote the sorted node index sequence as follows: , and ; all nodes in the region are sorted by reliability priority to prepare for selecting master nodes and backup nodes;

[0219] Set the first number of master nodes in the region as:

[0220] ;

[0221] Select the node with the minimum total cost in the region as the master node to ensure that the most reliable node in the region participates in sampling;

[0222] Set the first number of backup nodes in the region as:

[0223] ;

[0224] If , the number of backup nodes is ;

[0225] Select the node immediately following the master node in the order as the backup node to quickly replace the master node in case of failure;

[0226] Assign a backup activation flag to all nodes in the set , initialized to 0; if , it is considered a master node, and remains 0; the flag indicates which backup nodes have been replaced as master nodes to avoid repeated replacement or conflicts.

[0227] This part sets the number of master nodes in the region (rounded up to ensure redundancy) and selects master nodes and backup nodes based on the total path cost from small to large, to build a master-backup mutual backup mechanism. Through the backup activation flag, the state of the backup node is dynamically tracked to avoid repeated switching or resource waste. This mechanism uses a cost-driven intelligent selection strategy to achieve reliability configuration and rapid replacement potential evaluation at the node level. It solves the problems of fixed configuration and response lag in traditional redundancy design, improves the autonomy and emergency response speed of the acquisition system, ensures that the acquisition task can still run continuously under local fault conditions, and guarantees the safety and uninterrupted operation of data in explosive places.

[0228] The sampling period is divided into multiple time slots, and each master node is allocated a specific sampling time slot, which specifically includes:

[0229] Set the sampling period as ; define the length of a complete sampling- uploading cycle;

[0230] and divide into one time slot, each time slot length is:

[0231] allocating specific upload time for each master node to ensure conflict-free and orderly communication;

[0232] for the first master node in the region , the candidate time slot number is defined in turn as:

[0233] ; wherein, is a floor function; each master node in the region is preliminarily allocated a time slot to avoid internal conflict, while achieving uniform distribution on the global time slot;

[0234] The global time slot occupation marker array is set as , initialized to all 0; used to mark the allocated time slot to prevent cross-region or cross-node time slot conflict;

[0235] According to the region number , the following operations are performed in turn:

[0236] S400, for each master node in the region , set the candidate time slot number ; get the preliminary time slot number of the node to try to allocate; S500, if

[0237] , let the allocation function , and ; directly allocate the selected time slot to the master node when there is no conflict; S600, if

[0238] , start from and sequentially increase, if increased to , let , check in turn until a time slot with a value of 0 is found, then assign the time slot to and mark ; if the candidate time slot is occupied, delay to the next free time slot to ensure no conflict; S700, until the time slot allocation of the

[0239] th master node is completed, continue with the next master node; Finally, the unique sampling time slot of each master node

[0240] in each cycle is , corresponding to the actual sampling time ​​The time slot number is converted into an actual time, indicating when the node samples and uploads data; at this time, the node collects a measurement value and sends it to the sink node, performs a specific sampling operation, and completes data collection and uploading; wherein, is a period index.

[0241] The sampling period is divided into equal-length time slots, and each master node is assigned a specific sampling time slot, which is a key link to ensure that node data uploading is collision-free and time-controllable. Through the candidate time slot mechanism, the global flag array, and the incremental search algorithm, the organic combination of conflict avoidance and resource balanced scheduling is achieved. This scheduling algorithm solves the channel conflict and data collision problems caused by the concentrated uploading of nodes in the same area, greatly improves the network transmission efficiency and data success rate, and guarantees the high throughput and low delay characteristics of the entire collection system. At the same time, its flexible scalability also adapts to dynamic changes such as node addition and deletion, ensuring long-term stable operation of the system.

[0242] The master node fault detection is performed during operation, and the standby node takes over to perform the sampling task when a fault occurs, specifically including:

[0243] The first period is set as the sampling period of the master node, and the second period is set as the sampling period of the standby node.

[0244] ;

[0245] wherein, is the sequence number of the node in the set of master nodes in its area; records whether the data of a master node is successfully uploaded in each sampling period, providing basic information for fault determination; is the set of master nodes of the first sub-area; is the reception flag of whether the sink node has received the data of the master node in the first sampling period; The fault counter of the first

[0246] node at the end of the first period is set as:

[0247] , ;

[0248] Records the number of recent consecutive packet losses of the node, which can be considered as a real fault when there are multiple packet losses;

[0249] The continuous fault determination threshold is set as ; when ​When the value is large: Improved "jitter resistance": Tolerates one or two occasional network jitters or packet losses, preventing immediate misjudgment as node failure and reducing the possibility of incorrect backup node replacement; Reduced backup node startup frequency: Switches only when a node experiences multiple consecutive failures, reducing operational intervention and wasted backup resources. When a node truly fails hardware, wait for continuous... A fault is only diagnosed when packets are lost in every cycle, delaying the switchover time. During this delay, if the node is a critical master node in the region, its data loss will directly lead to the inability to fully monitor the region for multiple consecutive cycles, potentially missing out on sudden emergencies. When the value is small: If packet loss occurs within a certain period, a fault is immediately identified, and a backup node is activated, ensuring rapid recovery of area monitoring; the response to network jitter or hardware failure is more timely, reducing the impact of data loss on the overall system stability. A single sporadic network packet loss (e.g., during channel peak hours or temporary interference) may be misjudged as a "node failure," triggering backup node replacement and causing unnecessary waste of backup resources; frequent activation / deactivation of backup nodes leads to greater time slot reallocation and communication overhead, affecting the stability of the entire network. This solution recommends using a smaller value. In explosion-hazardous locations, high real-time performance is crucial. If the primary node fails, a backup node must take over as quickly as possible to ensure no more hazard signals are missed. While occasional packet loss in general wireless transmissions usually occurs within a single time slot or period, packet loss across two consecutive periods is more likely to indicate a genuine fault (internal hardware failure or prolonged obstruction). Therefore, [the appropriate node is selected]. It can avoid misjudgment in most cases of occasional packet loss, and can complete the switchover in just two cycles after a real node failure, and can also respond quickly to emergencies.

[0250] when At that time, node It was identified as a malfunction;

[0251] If in the cycle A master node exists at the end. satisfy If a master node experiences packet loss for two consecutive cycles, the following replacement operation will be performed. If a master node experiences packet loss for two consecutive cycles, the fault handling logic will be triggered:

[0252] S810, will Remove from the set of master nodes, specifically: Remove nodes that have been identified as faulty from the master node list and stop their subsequent sampling and uploading;

[0253] S820, in the set of backup nodes In the process, search for the first inactive node in ascending order of index. That is, satisfying ;

[0254] S830, the node joins the master node set, in particular:

[0255] , ;

[0256] Quickly select the optimal backup node to replace the failed node in the area to ensure the continuity of area monitoring;

[0257] S840, for the newly joined node allocate time slots:

[0258] S841, let the candidate time slot number be ; generate a priority time slot number for the new replacement node to maintain consistency with the original logic;

[0259] S842, if ,

[0260] then let , and update ;

[0261] wherein, is the backup node decrement the ranking in by 1;

[0262] S843, otherwise, sequentially increase to the next free time slot until an idle time slot is found and marked;

[0263] Avoid time slot conflicts between the replacement node and existing master nodes to ensure its unique sampling time slot after the current period;

[0264] S850, initialize the failure counter of the new replacement node : ; clear its historical packet loss record and treat it as a brand new master node for the next round of monitoring;

[0265] If there is no available backup node, the area only retains the remaining master nodes to continue the job.

[0266] Through periodic reception flags and continuous packet loss determination mechanisms, master node failure detection is achieved, combined with backup node activation logic and dynamic time slot allocation, which realizes a closed-loop processing of failure replacement. Especially when the sampling mission-critical node fails, the backup node can quickly complete the replacement to ensure uninterrupted collection. This scheme balances system stability and rapid response capability, effectively solving the problem of regional data loss and overall system failure risk caused by node failure. It is suitable for industrial safety places with extremely high reliability and real-time requirements, greatly enhancing the system's anti-interference, anti-failure and self-recovery capabilities.

[0267] The collected data is sent by the master node to the aggregation node, and data integrity check is performed after the end of the period, specifically including:

[0268] For the first Each master node in the period Send measurement value To set the reception of the aggregation node:

[0269] ;

[0270] Construct the first Periodic network sampling data matrix :

[0271] ;

[0272] ; wherein, is a sampling result record three tuple of the master node in the first sampling period;

[0273] Store all sampling entries in this period to facilitate unified verification and record;

[0274] Set the data integrity index of the first period in the first region as:

[0275] ;

[0276] Used for regional judgment of data integrity, providing partition alarm basis for maintenance personnel;

[0277] Set the data integrity index of the first period in the network as:

[0278] ; only when all , , otherwise The value of is 0; if the integrity flags of all sub-regions are 1, the network data integrity is 1, otherwise 0, which is convenient for overall monitoring overview;

[0279] The aggregation node synchronously stores the following information after the end of each period and can upload it to the monitoring center on demand: , ; archive the key information by period to facilitate subsequent manual analysis, fault positioning and operation and maintenance decision.

[0280] Through the construction of the periodic sampling data matrix, the calculation of the regional integrity index and the whole network integrity index, the data integrity analysis mechanism from local to global is proposed. The mechanism can quickly identify data missing, node abnormal or regional sampling failure and other problems at the end of the cycle, and record it to the monitoring system for back-end analysis. It solves the problem of the existing system "uploading is qualified" and makes the system have self-verification ability, which can provide high-quality, traceable complete data basis for upper-level decision-making. It improves the system operation transparency, supervision intensity and maintenance response efficiency.

[0281] It should be noted that the relational terms herein such as first and second and the like are used solely to distinguish one from another entity or action without necessarily requiring or implying any actual relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0282] The above description is only the preferred embodiment of the present application, and it should be pointed out that for those skilled in the art, without departing from the technical principles of the present application, a number of improvements and refinements can be made, which should be considered as the protection scope of the present application.

Claims

1. A data collection method for an explosion-protected device in an explosion- hazardous area, characterized in that, The method comprises the following steps: A three-dimensional Cartesian coordinate system of the scene is established, and three-dimensional coordinates of each sensor node are obtained; The three-dimensional Cartesian coordinate system of the scene is established, and the three-dimensional coordinates of each sensor node are obtained, specifically comprising: The explosion scene is a rectangular space, the bottom surface of which is a rectangle including two long sides and two short sides, and the length of the long side is greater than that of the short side; The center point of the bottom surface is obtained as the origin of the coordinate system, denoted as ; A ray parallel to the long side of the base is drawn through the origin, and the ray intersects the two short sides of the base. The ray parallel to the long side of the base is taken as the axis, and the direction of the ray is taken as the positive direction of the axis. A ray parallel to the short side of the base plane is made through the origin, and the ray intersects the two long sides of the base plane, respectively. The ray parallel to the short side of the base plane is taken as the axis, and the direction of the ray is the positive direction of the axis. A vertical ray is made through the origin, the direction of the ray is vertical upward, the ray intersects the upper and lower two bottom surfaces of the rectangular space respectively, taking the vertical ray to the bottom surface as the coordinate system axis, the direction of the ray is the positive direction of the axis; Deploying a plurality of sensor nodes in a coordinate system , numbered as , each node is represented by a three-dimensional coordinate ; wherein is the coordinate value of the node on the axis; is the coordinate value of the node on the axis; is the coordinate value of the node on the axis; The sensors are all of the same model; A convergence node is set up, numbered ; The three-dimensional coordinates of the sink node are denoted as ; wherein, is the coordinate value of the sink node on the axis; is the coordinate value of the sink node on the axis; is the coordinate value of the sink node on the axis; The scene is divided into multiple sub-regions according to the spatial structure, and each node is attributed to a corresponding region; The scene is divided into multiple sub-regions according to the spatial structure, and each node is attributed to a corresponding region, specifically comprising: The explosion site is divided into non-overlapping cuboid sub-regions with boundaries parallel to the coordinate axes in the coordinate system, numbered ; No. The region is represented by the following six parameters, specifically: ; wherein, and are the minimum and maximum coordinates of the first region on the x-axis, respectively; and are the minimum and maximum coordinates of the first region on the y-axis, respectively; and are the minimum and maximum coordinates of the second region on the x-axis, respectively; and are the minimum and maximum coordinates of the second region on the y-axis, respectively; and are the minimum and maximum coordinates of the third region on the x-axis, respectively; and are the minimum and maximum coordinates of the third region on the y-axis, respectively; The node attribution function is set as: ； When the node coordinates are on the multi-region overlapping boundary, the region with the smallest number To belong; Set the first The set of nodes within the region is: , ; wherein, is the number of regional nodes; is the number of elements of the set The failure probability of the sensor node and the failure probability of the communication link are calculated based on historical monitoring data; The failure probability of the sensor node and the failure probability of the communication link are calculated based on historical monitoring data, specifically comprising: For the first The sensor performs continuous monitoring, and the total monitoring sampling period is 1000. During the monitoring period, record the first... The number of missing data points for node number is denoted as . ; For each sensor node , let its single point failure rate be , and satisfy ; The sink node is denoted ; For each pair of nodes , the three-dimensional Euclidean distance between the two points is calculated: ， ； The maximum distance of communication of the sensor node is denoted as The maximum distance of communication of the sensor node is denoted as as the effective distance threshold of inter-node communication; A node may communicate with a node if and only if , otherwise there is no communication link. If , the communication link between the setting node and the node fails with a probability of in one transmission process. The communication reachability indication is set as: ; wherein, represents a node communicable with a node ; represents a node incommunicable with a node ; ; The network topology structure is constructed based on the node communicability judgment, and the optimal reliable path of each node to the sink node is determined according to the cost function of the node and the link; The network topology structure is constructed based on the node communicability judgment, and the optimal reliable path of each node to the sink node is determined according to the cost function of the node and the link, specifically comprising: The set of all nodes is set to ; wherein the number 0 represents the sink node, the numbers 1 to represent the sensor nodes; The edge set is set to ; wherein, represents a node with an available link between the node in a communication cycle; Constructing an undirected graph is ; For each node , set the node cost to ; where ; For each edge , set the edge cost to , ; For any node, let a simple path from the node to the sink node be: ; wherein, denotes the th node on the path; and all are distinct; is the number of intermediate nodes on the path from node to the sink node. The total reliability of the path is set as: ； The total cost of the path is set as: ; wherein, is the edge cost of the edge between the th node; is the node cost of the th node; The optimal path search algorithm is set, specifically comprising: S100, initialization: S101、set groups current minimum path cost from the node to the sink node: ， ； S102、Set the predecessor array , initialized to undefined; S103, set the optimal cost node set which has been fixed ; S200, iteration process: In Non-vacuously, repeat the following operations: S201, Order , will node Merge into an existing fixed set ;like If the condition is met, the loop exits and the algorithm terminates; otherwise, it continues. S202, for each and There are connected edges and none have been added. nodes Perform a relaxation update: Computing by Total candidate cost to reach the sink node as a relay ; If , then update , ; Step S200 is repeated until all nodes have joined or the remaining nodes are unreachable; S300, path backtracking and result output: For each number , if , then by array from backtracking to the convergence node, the optimal path ; at the same time, is the total cost of the optimal path; If , then the node has no reachable path to the sink node; Redundancy planning is performed for each sub-region, and the master node and the backup node are selected based on the path cost; A sampling period is divided into multiple time slots, and each master node is allocated a specific sampling time slot; Master node failure detection is performed during operation, and the backup node takes over the sampling task when a failure occurs; The collected data is transmitted by the master node to the sink node, and data integrity verification is performed after the period ends.

2. The data collection method for an explosion-proof apparatus in an explosion- hazardous location according to claim 1, wherein, The redundancy planning is performed for each sub-region, and the master node and the backup node are selected based on the path cost, specifically comprising: For the first region, the number of master nodes is set as ; wherein is a ceiling function. In the first region, all nodes are arranged in ascending order according to the path cost , and the node index sequence after sorting is recorded as , and ; Set the first The set of master nodes for the region is: ； The first setting The set of backup nodes for the region is: ； If then the number of spare nodes ; For sets All nodes are assigned a spare activation flag. Initialize to 0; where, if If it is, then it is considered the master node. Keep it at 0.

3. The data collection method for an explosion-proof apparatus in an explosion- hazardous location according to claim 2, wherein, The sampling period is divided into multiple time slots, and each master node is allocated a specific sampling time slot, specifically comprising: The sampling period is set to ; and will be divided into slots, each of length: ； For the first The first in the region One master node Candidate time slot numbers are defined sequentially as follows: ; wherein is a floor function; Setting the global slot occupancy flag array to , all initialized to 0; by region number the following operations are performed in this order: S400、to the region each master node , set candidate time slot number ; S500、if then let the allocation function and set ; S600、if , from start sequentially increasing, if increased to then let , check in turn until the time slot with value 0 is found, then assign the time slot to and mark ; S700, up to the... Once the time slots for the first master node are allocated, proceed to the next master node. Eventually, each master node At each cycle The unique sampling slot within the cycle Corresponds to the actual sampling time At this time, the node Collects a measurement value and sends it to the sink node; wherein, Is the cycle index.

4. The data collection method for an explosion-proof apparatus in an explosion- hazardous location according to claim 3, wherein, The master node failure detection is performed during operation, and the backup node takes over the sampling task when a failure occurs, specifically comprising: The first periodic node receiving flag is set as: ； wherein, is a node a sequence number in its set of regional master nodes; is a set of master nodes for the th sub-region; is a reception flag of whether the sink node has received the data of the master node in the th sampling period. The first node is set to have a fault counter of zero at the beginning of the first period. ， ； Let the continuous fault determination threshold value be ; When the node is identified as faulty; If at the end of the cycle there is a master node satisfying then the following replacement operation is performed: S810, the eliminate from the master node set, specifically: ; S820、In the standby node set from the smallest to the largest index to find the first node that is not activated , that is, satisfies ; S830, the node join the master node set, specifically: ， ； S840, for the newly joined node Allocated time slots: S841, Let the candidate time slot number be... ; S842, if , Let and update ; wherein, is a backup node in ranked one less in S843, otherwise, sequentially increase to the next free time slot until a free time slot is found and marked; S850, for the new replacement node Fault counter initialization: ; If If there are no spare nodes available in the zone, then the zone Only the remaining primary nodes remain to continue the job.

5. The data collection method for an explosion-proof apparatus in an explosion- hazardous location according to claim 4, wherein, The collected data is transmitted by the master node to the sink node, and data integrity verification is performed after the period ends, specifically comprising: For the first each primary node at time sending measurement values to gather node reception settings: ； Constructing a first Periodic full network sampled data matrix : ； ; wherein, is the th sampling result record triplet of the master node in the th sampling period. The first The first The periodic data integrity index is: ； Setting the whole network first Periodic data integrity index is: ; if and only if all , , otherwise the value of 0; The sink node stores and synchronizes the following information at the end of each cycle and can upload it to the monitoring center on demand: , .

Images