An Adaptive Topology Control Method for Monitoring Nodes in a Confined Space

By using an adaptive topology control method, the problem of monitoring blind spots caused by fixed deployment of monitoring nodes in confined space operations is solved, realizing adaptive layout and dynamic coverage of nodes, ensuring full-area coverage and continuous network operation capability.

CN123113299APending Publication Date: 2026-11-17河北云酷科技有限公司
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611470671.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-09-21
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

In existing technologies, monitoring nodes in confined space work environments are deployed at fixed locations and cannot be adjusted according to the actual shape, resulting in a lack of effective monitoring coverage in areas far from the installation point or in areas that are obstructed, which poses a safety hazard.

Method used

An adaptive topology control method is adopted, which realizes the adaptive layout and dynamic coverage of monitoring nodes through on-site survey zoning, virtual force adjustment and self-organizing network technology. The working roles are dynamically allocated based on the node energy status and distribution density, and a topology maintenance mechanism of responsive local reconstruction and preventive reconstruction is established.

Benefits of technology

It achieves a matching layout between the monitoring network and the spatial form, eliminates monitoring blind spots, ensures full-area coverage and continuous network operation capability, and improves the continuity and reliability of data transmission.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123113299A_ABST
    Figure CN123113299A_ABST
Patent Text Reader

Abstract

This invention relates to the field of safety monitoring for confined space operations, specifically disclosing an adaptive topology control method for monitoring nodes in confined space operations. The method includes conducting on-site surveys of the target confined space and dividing the space into several sub-regions based on the survey results; configuring corresponding initial topology types for each sub-region and placing monitoring nodes; each monitoring node undergoing adaptive positional fine-tuning through virtual force adjustment until the full space coverage reaches a preset requirement; all monitoring nodes are then activated to autonomously determine their operating mode based on the energy state and distribution density of their one-hop neighbor nodes and generate a global topology; and each monitoring node collects environmental parameter data and transmits it to the central control system through the global topology. This invention achieves adaptive deployment and dynamic topology control of monitoring nodes in complex and irregular spaces through the coordinated operation of zoning planning, virtual force-driven positional fine-tuning, and energy-sensing topology generation.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of safety monitoring for confined space operations, specifically an adaptive topology control method for monitoring nodes in confined space operations. Background Technology

[0002] Confined space work environments are generally characterized by high degree of enclosure, complex internal structure, and high risk of harmful gas accumulation. In scenarios such as municipal pipelines, mine tunnels, and inside chemical equipment, the safety monitoring of workers has always been a key focus in the field of industrial safety. Traditionally, environmental monitoring in confined spaces mainly relies on installing sensors at fixed locations. These locations are typically chosen at the space entrance, near known hazards, or locations where problems have been concentrated in historical monitoring. The collected data is then aggregated to a central control system via wired or wireless methods.

[0003] However, the internal structure of confined spaces is often highly irregular. Underground pipelines may extend for thousands of meters and have numerous bends and branches; chemical reactors contain obstacles such as agitators and baffles; and mine tunnels extend continuously with the progress of mining. These spatial variations create complex obstructions and limitations on the sensing range of sensors. Once fixed-point sensors are installed, their location and number no longer adjust to the actual spatial structure, resulting in areas far from the installation points or parts blocked by obstacles lacking effective monitoring coverage. If harmful gases accumulate or oxygen deficiency occurs in these areas during operations, it will not be detected in time, posing a direct safety threat to workers entering these areas. Therefore, how to adapt the layout of monitoring nodes to the actual shape of confined spaces and achieve effective coverage of the entire area has become a pressing technical problem to be solved in this field. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides an adaptive topology control method for monitoring nodes in confined spaces. This method solves the problem in the prior art where monitoring nodes are deployed at fixed locations and cannot adjust their positions according to the actual shape of the confined space, resulting in a lack of effective monitoring coverage in areas far from the installation point or that are obstructed.

[0005] This invention provides an adaptive topology control method for monitoring nodes in confined spaces, comprising the following steps:

[0006] Step 1: Conduct on-site survey of the target confined space to obtain information on the spatial morphology, bifurcation node location, and occlusion area distribution of the confined space. Based on the survey results, divide the confined space into several sub-regions and configure the corresponding initial topology type for each sub-region.

[0007] Step 2: Place monitoring nodes according to the initial topology of each sub-region. After the nodes are placed, each monitoring node is finely adjusted in position through a virtual force adaptive adjustment process. The virtual force includes the virtual repulsive force between adjacent monitoring nodes and the virtual attractive force of the covered void area on the monitoring node. Each monitoring node moves gradually according to the direction of the virtual resultant force until the full space coverage reaches the preset requirement.

[0008] Step 3: Start all monitoring nodes. Each monitoring node autonomously determines its own working mode based on the energy state information and distribution density information of its surrounding one-hop neighbor nodes, and completes local clustering based on the working mode to generate a global topology.

[0009] Step four: Each monitoring node collects environmental parameter data within a limited space and transmits the environmental parameter data to the central control system through the global topology.

[0010] Preferably, in step two, the virtual repulsive force acts between any two adjacent monitoring nodes whose actual distance is less than the preset optimal spacing, driving the two monitoring nodes to move away from each other;

[0011] The virtual attraction acts between the monitoring node and a spatial area not covered by the perception range of any node, driving the monitoring node to move closer to that area.

[0012] The virtual resultant force is determined by the vector superposition of the virtual repulsive force and the virtual attractive force.

[0013] Preferably, in step two, the method for determining whether the full space coverage rate meets the preset requirement is as follows:

[0014] The finite space is discretized into several grid cells. The number of grid cells covered by the sensing range of at least one monitoring node is counted. The ratio of the number of covered grid cells to the total number of grid cells is calculated. When the ratio is greater than or equal to a preset coverage threshold, the virtual force adaptive adjustment process is determined to be complete.

[0015] Preferably, in step three, each monitoring node autonomously determines its own operating mode based on the energy state information and distribution density information of its surrounding one-hop neighbor nodes, specifically as follows:

[0016] Each monitoring node obtains its own remaining energy and the average remaining energy of its one-hop neighbor nodes, compares its own remaining energy with the average remaining energy, and determines its own energy level.

[0017] Each monitoring node obtains the number of one-hop neighbor nodes and calculates the distribution density of its local area in combination with its own communication coverage area;

[0018] For monitoring nodes at high energy levels and whose local distribution density is below a preset lower limit, the routing backbone working mode is selected;

[0019] Monitoring nodes located at low energy levels or with a local distribution density exceeding a preset upper limit should select the terminal acquisition mode; other monitoring nodes should select the relay forwarding mode.

[0020] Preferably, after the global topology is generated, each monitoring node continuously monitors the signal quality of the communication link. When any node fails or the signal quality of any communication link is lower than a preset quality standard, a local topology reconstruction is triggered.

[0021] The scope of the local topology reconstruction is limited to the one-hop neighbors of the failed node or abnormal link, and after reconstruction, it is re-integrated into the global topology structure.

[0022] Preferably, in step one, the initial topology type includes linear chain topology and branching topology;

[0023] The linear chain topology is configured for sub-regions with a long and narrow spatial shape, and the branching topology is configured for sub-regions with branching nodes.

[0024] Preferably, in step one, each of the sub-regions is configured with a corresponding coverage redundancy parameter, wherein the coverage redundancy is used to characterize the degree of overlap of the sensing range of the monitoring nodes within the sub-region;

[0025] For sub-regions containing high-risk key monitoring areas marked in the survey results, a first coverage redundancy is configured, and for sub-regions containing non-high-risk key monitoring areas, a second coverage redundancy is configured, with the first coverage redundancy being greater than the second coverage redundancy.

[0026] Preferably, in step three, each monitoring node exchanges topology information only with a one-hop neighbor node, and each monitoring node independently decides its own transmission power level based on the number and location information of the one-hop neighbor nodes, selecting the lowest transmission power level while maintaining connectivity with at least one neighbor node.

[0027] Preferably, the monitoring node is a modular node, including independently replaceable sensor modules, communication modules, and power supply modules; the sensor modules are equipped with different types and accuracies of sensors according to monitoring requirements; for sub-areas with multi-directional branch extension requirements, branch relay nodes or three-way switching nodes are configured, the branch relay nodes have multiple communication interfaces, and can simultaneously establish links with neighboring nodes in three or more directions to realize multi-level branch expansion in space; for sub-areas with a narrow and long orientation that require long-distance coverage, multi-hop communication between adjacent nodes extends step by step along the length direction, and new nodes can automatically discover and access the existing topology to realize the infinite extension of the monitoring network in the length direction.

[0028] Preferably, the monitoring node also integrates a positioning module to obtain its own or the location information of personnel / equipment carrying positioning tags in real time; the location information is transmitted to the central control system through a global topology structure to realize the real-time positioning and trajectory tracking of workers in a confined space; the environmental parameter data collected by each monitoring node is associated and packaged with the location information at the corresponding time to form monitoring data with spatial coordinates.

[0029] Compared with the prior art, the present invention has the following beneficial effects:

[0030] Through on-site surveys and zoning planning, appropriate initial topology types are configured for different sub-regions based on the actual shape and bifurcation structure of the limited space. This ensures that the layout of the monitoring network matches the spatial geometric characteristics from the very beginning of deployment, realizing the transformation from fixed-point deployment to shape-adaptive deployment. This eliminates monitoring blind spots caused by spatial irregularities from the source and lays a structural foundation for subsequent refined coverage control.

[0031] This invention introduces a virtual force-driven adaptive node position adjustment mechanism, enabling nodes to autonomously detect coverage gaps and move to fill them after deployment based on the distribution of their neighbors. Simultaneously, it dynamically allocates three types of working roles—routing backbone, relay forwarding, and terminal data collection—based on node energy status and local density information. This achieves coordinated optimization of node spatial distribution and communication load, balancing network energy consumption while ensuring full-area coverage integrity, and effectively extending the continuous operation capability of the monitoring system.

[0032] This invention establishes a topology maintenance mechanism that integrates reactive local reconfiguration and preventative reconfiguration. This mechanism enables the network to respond quickly to abnormal situations such as unexpected node failures, signal quality degradation, or premature energy depletion. It allows for both rapid local repair after a fault occurs and proactive adjustment of the network structure based on energy decay trends before a fault actually occurs. This upgrades topology control from passive response to proactive prevention, ensuring the continuity and reliability of data transmission during long-term operation. Attached Figure Description

[0033] Figure 1 This is a simplified schematic diagram illustrating the process of monitoring node deployment and topology generation in Embodiments 1 and 2 of the present invention;

[0034] Figure 2 This is a simplified flowchart illustrating the preventive topology reconfiguration mechanism in Embodiment 3 of the present invention.

[0035] Figure 3 This is a schematic diagram of the modular node configuration in Embodiment 4 of the present invention. Detailed Implementation

[0036] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0037] Example 1: This example uses an underground utility tunnel with a total length of approximately 120 meters and two internal branches as an application scenario to explain in detail the deployment of monitoring nodes and the topology control process.

[0038] I. On-site survey and zoning

[0039] Construction workers first conducted a comprehensive survey of the utility tunnel. The main tunnel is long and narrow, approximately 2.5 meters wide and 2.8 meters high. At approximately 50 meters in the middle and 95 meters at the end, two branch tunnels extend to the left and right respectively, with lengths of approximately 20 meters and 15 meters. During the survey, workers used laser rangefinders and handheld 3D scanning equipment to obtain information on the spatial morphology and branching points of the tunnel, and marked obstructions within the tunnel, primarily including three areas obstructed by ventilation ducts and equipment supports. Simultaneously, based on historical maintenance records and on-site gas detection data, the area from approximately 30 to 55 meters in the middle section of the tunnel was designated as a high-risk monitoring area, where abnormal methane concentrations had been detected multiple times.

[0040] After the survey was completed, the utility tunnel was divided into several sub-regions based on its spatial morphology: the section from 0 to 50 meters of the main tunnel was the first linear segment sub-region, the section from 50 to 95 meters was the second linear segment sub-region, and the section from 95 to 120 meters was the third linear segment sub-region; the two bifurcation nodes were respectively designated as the first branch node sub-region and the second branch node sub-region. For the linear segment sub-regions, a linear chain topology was configured; for the branch node sub-regions, a branch topology was configured. Each sub-region was configured with corresponding coverage redundancy parameters. For the first linear segment sub-region, which contains high-risk key monitoring areas, a first coverage redundancy was configured, requiring that any location within this area be simultaneously covered by the sensing range of at least three monitoring nodes; for sub-regions not containing high-risk key monitoring areas, a second coverage redundancy was configured, requiring that any location be covered by the sensing range of at least one monitoring node. Coverage redundancy characterizes the degree of overlap in the sensing ranges of monitoring nodes within a sub-region; the higher the redundancy, the more layers of simultaneous coverage by multiple nodes in the same area.

[0041] II. Node Layout and Adaptive Adjustment of Virtual Force

[0042] The monitoring nodes used in this embodiment are cylindrical modular nodes, approximately 12 cm in diameter and 15 cm in height, with a micro-wheel drive mechanism at the bottom, allowing them to move autonomously on the bottom surface of the utility tunnel. Each node integrates multiple gas sensors, including methane, carbon monoxide, hydrogen sulfide, and oxygen sensors, as well as temperature and humidity sensors, a wireless communication module, and a power supply module. The wireless communication module is based on the ZigBee protocol, with a communication radius of approximately 15 meters; the power supply module is a built-in rechargeable lithium battery with a nominal capacity of 10,000 mAh.

[0043] Following the pre-defined topology, the construction team first roughly placed the nodes in their planned locations. In this embodiment, a total of 18 monitoring nodes were deployed: 12 on the main linear channel and 3 on each of the two branch channels. After power-on, each node automatically entered the virtual force adaptive adjustment phase.

[0044] The virtual force between nodes is determined as follows. For any two adjacent monitoring nodes... and If the actual distance between the two is less than the preset optimal spacing This creates a virtual repulsive force between the two, causing them to move away from each other. The magnitude of this repulsive force is proportional to the degree to which the distance between them deviates from the optimal distance. In this embodiment, based on the width of the pipe gallery and the node sensing radius, the effective sensing radius of the gas sensor at a single node is approximately 3 meters, and the preset optimal distance... Take 6 meters.

[0045] Monitoring nodes Received from nodes Virtual repulsive force Determine by the following formula:

[0046]

[0047] in, For monitoring nodes With monitoring nodes The current distance between them The repulsive force coefficient is taken as in this embodiment. , To monitor nodes Point to monitoring node The unit vector. If ,but That is, no repulsive force is generated between the two nodes.

[0048] Meanwhile, each monitoring node continuously detects coverage holes around itself. The detection method for coverage holes is as follows: Each node obtains the location information of its one-hop neighbor node, forms a coverage circle with its own location as the center and its sensing radius as the radius, and performs geometric calculations with the coverage circles of its one-hop neighbors. The gap between these coverage circles is the coverage hole region, and the boundary point of the gap region is the coverage hole boundary point. When a coverage hole region is detected, it exerts a virtual attraction on nearby monitoring nodes, driving them to move towards the hole region to fill the coverage blind spot.

[0049] Monitoring nodes Virtual attraction Determine by the following formula:

[0050]

[0051] in, For monitoring nodes With the The distance between the boundary points of the covered voids The attractiveness coefficient is taken in this embodiment. , To monitor nodes Pointing to the A unit vector covering the boundary points of the cavity. For monitoring nodes Total number of detected boundary points of covered voids.

[0052] Monitoring nodes Virtual combined force It is a vector superposition of repulsive and attractive forces:

[0053]

[0054] in, For monitoring nodes The set of one-hop neighbor nodes. Each monitoring node moves in the direction of the virtual resultant force it receives, with each movement taking a step size of... Take 0.2 meters. After the movement is completed, the resultant force of each node is recalculated and the node is moved again, and this process is repeated iteratively.

[0055] During the virtual force adjustment process, each node performs the following judgment after each round of movement: the system records the cumulative movement distance increment and communication link establishment time increment of each node after each round of adjustment, and judges whether the benefit of continued adjustment is approaching saturation based on the improvement rate change between adjacent rounds. Specifically, when the improvement rate is lower than the preset slope threshold for three consecutive rounds, it is determined that the current topology has reached stability, and the virtual force adjustment is terminated early. This mechanism avoids continuing to consume node energy in iteration rounds with extremely low adjustment benefits.

[0056] In this embodiment, the virtual force adjustment process iterated for approximately 15 rounds. After each iteration, the system calculated the current coverage rate using a grid discretization method: the utility tunnel space was discretized into 0.5m × 0.5m grid cells, and the number of grid cells covered by the sensing range of at least one monitoring node was counted. Before the first iteration, the initial coverage rate was approximately 73%; after 15 iterations, the coverage rate increased to 96.8%, meeting the preset coverage threshold, which in this embodiment was 95%, at which point the virtual force adjustment process terminated. At this point, the nodes formed a roughly uniform spatial distribution within the utility tunnel, and the coverage redundancy of the high-risk key monitoring areas met the triple coverage requirement.

[0057] III. Ad hoc networks and topology generation

[0058] To facilitate understanding of the above deployment and networking process, Figure 1 A simplified diagram illustrating the complete process of monitoring nodes from on-site survey to data upload is provided. Combined with... Figure 1 The process shown below will be further explained in detail regarding the virtual force adaptive adjustment and self-organizing network process.

[0059] After the virtual energy adjustment is completed, all monitoring nodes enter the networking phase. Each node first obtains its own remaining energy and the average remaining energy of its one-hop neighbor nodes to determine its energy level. In this embodiment, the energy level classification coefficient... Take 1.2, Take 0.6. If the remaining energy of a node is greater than 1.2 times the average remaining energy of its one-hop neighbors, then the node is at a high energy level; if the remaining energy is less than 0.6 times the average remaining energy of its one-hop neighbors, then the node is at a low energy level; if it is in between, then the node is at a medium energy level.

[0060] Simultaneously, each node obtains the number of one-hop neighbor nodes and calculates the distribution density of its local area based on its own communication coverage area. The distribution density is calculated as the ratio of the number of one-hop neighbor nodes to its own communication coverage area. Monitoring nodes at high energy levels and with a local area distribution density below a preset lower limit select the routing backbone working mode to undertake data aggregation and forwarding tasks; monitoring nodes at low energy levels or with a local area distribution density above a preset upper limit select the terminal acquisition working mode, only responsible for data acquisition and reporting, and not participating in data forwarding; the remaining monitoring nodes select the relay forwarding working mode, both acquiring data and forwarding data from neighboring nodes. In this embodiment, the preset lower limit of distribution density is 0.02 nodes per square meter, and the preset upper limit of distribution density is 0.08 nodes per square meter.

[0061] In this embodiment, 4 out of the 18 nodes are ultimately identified as routing backbone nodes, mainly distributed near high-risk areas and branch nodes; 10 nodes are relay forwarding nodes; and 4 nodes are terminal acquisition nodes. Each node completes local clustering according to its own working mode. Each routing backbone node and several relay nodes and terminal nodes within its one-hop range form a cluster, ultimately generating a global topology structure covering the entire utility tunnel.

[0062] The node transmit power decision-making method is as follows: In this embodiment, the node transmit power is divided into 8 levels, ranging from -18dBm to 4dBm, with increments of approximately 3dBm. Each node only exchanges topology information with one-hop neighbor nodes and does not acquire information about nodes two hops or more away. Each node independently decides its own transmit power level based on the number and location information of its one-hop neighbor nodes. Specifically, the node first sends a probe signal at the lowest power level. If it receives an acknowledgment response from at least one neighbor node, it maintains that power level; if it does not receive any response, it gradually increases the power until it receives at least one acknowledgment response, and then uses that power level as its current operating power. This decision-making process is re-executed whenever the topology information of each node changes. Through this method, each node selects the lowest transmit power level while maintaining connectivity with at least one neighbor node, effectively reducing overall energy consumption.

[0063] IV. Data Transmission

[0064] After the global topology is generated, each monitoring node collects environmental parameter data such as methane concentration, carbon monoxide concentration, oxygen content, temperature, and humidity within the utility tunnel according to a preset sampling period (30 seconds in this embodiment). The data is then relayed to the central control system at the tunnel entrance and exit via the generated global topology. The central control system receives the data and stores, displays, and performs anomaly analysis. During the 72-hour trial run, the system achieved a data acquisition integrity rate of 99.7%, with no data loss or communication interruptions.

[0065] Example 2: This example uses a complex internal reactor in a chemical plant as an application scenario. The reactor is a vertical cylindrical structure with an inner diameter of approximately 4 meters and a height of approximately 6 meters. It contains multiple layers of stirring blades and baffles, resulting in a complex spatial structure and significant obstruction. Traditional fixed-point monitoring methods are prone to creating blind spots in this scenario.

[0066] I. Survey and Zoning

[0067] Construction workers entered the reactor through a manhole on top to conduct a survey. Replacement ventilation and safety checks had been completed before the operation. The survey obtained three-dimensional spatial morphology information of the reactor's interior, identifying the locations of the three layers of agitator blades as the main shielding areas, situated at heights of 1.5 meters, 3.0 meters, and 4.5 meters from the bottom of the reactor, respectively. Since the reactor's interior lacks branching structures and is a single columnar space, it was divided into a sub-region with a branched topology. This involved forming a vertical backbone link, with nodes at each height level serving as branches connecting to the backbone node.

[0068] Based on the internal structure and process hazard level of the reactor, the area from the bottom of the reactor to a height of 1.5 meters, i.e. the main area of ​​the reactants, and the gas phase space at the top of the reactor, i.e. the area where flammable gases are easily accumulated, are marked as high-risk key monitoring areas and configured with a high degree of coverage redundancy.

[0069] II. Node Layout and Adaptive Adjustment

[0070] This embodiment deploys a total of 12 monitoring nodes. Unlike Embodiment 1, the internal space of the reactor is narrow and contains obstacles such as stirring blades, preventing the nodes from moving freely on the ground. Therefore, the nodes in this embodiment use magnetic mounting bases, allowing them to adhere to the inner wall of the reactor or the stirring shaft support. It should be noted that the fine-tuning of the monitoring node positions in this invention can be accomplished autonomously by the node's built-in drive mechanism or assisted by on-site personnel based on the virtual resultant force direction information fed back by the node. Both methods are specific implementations of "gradually moving according to the direction of the virtual resultant force."

[0071] After power-on, the nodes enter the virtual force calculation phase. Each node obtains the location information of its one-hop neighbor nodes through the communication module, calculates virtual repulsive and virtual attractive forces, but the nodes do not move automatically. Instead, they feed back the calculated virtual resultant force direction and suggested movement distance to the construction personnel via indicator lights or wireless means. The construction personnel manually adjust the node positions according to the feedback information, recalculating after each adjustment until the full space coverage reaches the preset requirements.

[0072] In this embodiment, since the reactor is a three-dimensional space, the virtual force calculation is performed in a three-dimensional coordinate system. A preset optimal spacing is used. Based on a node sensing radius of approximately 2.5 meters, a radius of 5 meters was chosen. After eight rounds of manual adjustments, the coverage rate increased from the initial 68% to 94.2%, reaching the preset threshold, which in this embodiment was 92%.

[0073] III. Ad hoc networks and data transmission

[0074] The networking phase is basically the same as in Example 1. Due to the small space inside the reactor and the strong reflection and attenuation of wireless signals by the metal walls, a slightly higher power margin was used when determining the transmission power level of the nodes in this example to ensure the stability of the communication link in a strong reflection environment.

[0075] After the global topology is generated, the system continuously monitors the signal quality of each communication link. In a simulation test, a relay node located near the second-layer agitator blades experienced a drop in communication signal quality below the preset quality standard due to vibrations caused by the blade rotation. Upon detecting this anomaly, the system automatically triggered a local topology reconfiguration. The reconfiguration was limited to the one-hop neighbors of the anomalous node, involving three adjacent nodes. The reconfiguration process was completed in approximately 2.3 seconds, without any perceptible impact on global data transmission. After reconfiguration, the new local topology was reintegrated into the global topology.

[0076] Example 3: This example uses an inclined roadway in a coal mine as an application scenario to illustrate the implementation process of the preventive topology reconfiguration mechanism.

[0077] The inclined tunnel is approximately 200 meters long with a slope of about 15 degrees. The tunnel cross-section is arched, about 3.5 meters wide, and about 3.0 meters high. Twenty-two monitoring nodes were set up along the tunnel, and the survey zoning, virtual force deployment, and networking were completed according to the method in Example 1.

[0078] I. Normal Operation Phase

[0079] After the system was put into operation, each monitoring node collected data such as methane concentration, carbon monoxide concentration, wind speed, temperature, and humidity according to a preset cycle, and uploaded the data to the ground-based central control system through the global topology. In the initial stage, the remaining energy of all nodes was above 85% of their nominal capacity, and the network operated stably.

[0080] II. Energy Decay Monitoring and Trend Prediction

[0081] During operation, the system continuously records the energy consumption rate and historical communication quality trends of each monitoring node. Due to the inclined shape of the roadway, nodes located at higher elevations (near the exit) and lower elevations (near the working face) experience different communication loads. Lower-elevation nodes bear more data relay and forwarding tasks, resulting in a relatively faster energy consumption rate.

[0082] On day 45, the system fitted an energy decay curve based on the historical energy consumption data of each node. For a backbone node A located at the bottom of the tunnel, its energy decay rate is approximately 1.8% per day. Based on the current rate, its remaining effective duration is estimated to be approximately 22 days. The system compares this prediction with a preset time threshold, which in this embodiment is 30 days. Finding that the remaining effective duration of 22 days is less than 30 days, a preventative topology reconfiguration is triggered.

[0083] III. Preventive Topology Reconstruction

[0084] Figure 2 A simplified flowchart illustrating the preventative topology reconfiguration mechanism is provided. (Combined with...) Figure 2 As shown in the process, in this embodiment, preventative refactoring is initiated before node A actually fails.

[0085] Preventive reconfiguration is initiated before node A actually fails. Based on the energy status and location information of each node within node A's one-hop neighbor range, the system re-evaluates the best backbone node candidates for that local area. Among node A's one-hop neighbors, node B is located approximately 5 meters above node A, has higher remaining energy than node A, and its communication coverage largely overlaps with node A's original coverage area. The system adjusts node B's operating mode from relay forwarding to routing backbone, appropriately adjusts node B's transmit power to cover the area originally covered by node A, and downgrades node A to a regular relay node to reduce its energy consumption.

[0086] The entire preventative reconfiguration process was completed while node A was still operating normally, without causing any data transmission interruption. Due to reduced energy consumption, node A continued operating for approximately 40 days before entering a low-power state. By this time, the system had already completed its topology adjustment, and network operation was unaffected. After another 20 days, the system triggered another round of preventative reconfiguration based on energy trend predictions, achieving a smooth topology transition. Preventative reconfiguration and local topology reconfiguration are independent topology maintenance mechanisms; the former is executed in advance based on predictions, while the latter is triggered by actual anomalies.

[0087] Explanation of preset parameters

[0088] The various preset parameters involved in the above embodiments, such as the preset optimal spacing repulsive force coefficient Attraction coefficient Energy level classification coefficient and The specific values ​​for thresholds such as coverage, time, and distribution density can be reasonably set based on factors such as the sensing radius, communication radius, spatial size, and monitoring accuracy requirements of the monitoring nodes in the actual application scenario. For example, in a relatively open space, the values ​​can be appropriately increased. To increase the spacing between nodes and reduce the number of nodes, the coverage threshold can be appropriately increased in scenarios with high monitoring accuracy requirements. For heterogeneous node mixed deployment scenarios with inconsistent sensing ranges, the preset optimal spacing can be determined based on the smaller sensing radius of the node pair. Adjustments to all parameters are within the scope of the technical solution of this invention and do not constitute a limitation on the scope of protection of this invention. Furthermore, the selection and configuration of each module of the modular node, the number and layout of communication interfaces of the three-way adapter node, and the specific positioning technology used by the positioning module (such as UWB, Bluetooth AoA, or Wi-Fi fingerprint positioning) can all be adaptively selected and configured according to the monitoring index requirements, spatial geometry, and positioning accuracy requirements of the actual scenario. These are all reasonable modifications to the technical solution of this invention and do not constitute a limitation on the scope of protection of this invention.

[0089] Example 4: This example uses a large underground utility tunnel as the application scenario. It is about 500 meters long and has 6 branch entrances along the way, which lead to different auxiliary equipment rooms and adjacent road sections. The internal space of the tunnel is narrow, with the narrowest point being only 1.8 meters wide, which puts forward high requirements for the miniaturization and flexible deployment of monitoring equipment.

[0090] I. Modular Node Configuration

[0091] This embodiment employs the modular monitoring node described in this invention. Each node consists of three main modules: a sensor module, a communication module, and a power supply module, combined via standard quick-connect interfaces. The sensor module can be flexibly selected according to the monitoring needs of different pipe gallery sections: methane and hydrogen sulfide sensors are selected for sections with dense gas pipelines, hydrogen sulfide and ammonia sensors are selected for sewage pipe sections, and oxygen and temperature / humidity sensors are added for sections with frequent personnel access. The communication module supports the ZigBee protocol and integrates a UWB positioning module. The power supply module uses a replaceable lithium battery with a nominal capacity of 15000mAh. All modules are electrically connected via standard interfaces, allowing for replacement or reconfiguration on-site within 3 minutes, realizing on-demand combination of monitoring functions.

[0092] II. Chain-like cascading infinite extension

[0093] The main tunnel has a long and narrow shape, stretching 500 meters from entrance to exit. Monitoring nodes are deployed along the main tunnel at 10-meter intervals, totaling 51 nodes. Each node extends towards the exit step-by-step via multi-hop relay communication between adjacent nodes. The furthest node from the exit is 500 meters away, requiring 49 relay hops. Data is still transmitted stably to the central control system at the exit in complete data packet format, with an average end-to-end transmission delay of approximately 2.8 seconds and a data integrity rate of 99.5%. As the tunnel continues to be excavated and extended, construction personnel only add new nodes at the same intervals along the extension section. Upon power-up, the new nodes automatically scan and discover existing network endpoints, automatically completing authentication and access via a handshake protocol. This eliminates the need to reconfigure the entire network, enabling unlimited extension of the monitoring network along its length.

[0094] III. Three-way adapter and multi-level branch expansion

[0095] At six branch points along the utility tunnel, T-junction nodes are installed. Each node has three independent communication interfaces, allowing simultaneous link establishment with two neighboring nodes in the main line direction and the first node in the branch direction. Communication links in each direction are maintained independently and do not interfere with each other. Taking the third branch as an example, this branch tunnel is 40 meters long and contains three monitoring nodes. The T-junction node maintains communication connections with nodes 10 meters before and after it in the main line direction, and establishes a link with the first node in the branch tunnel. Nodes within the branch tunnel continue to cascade and relay each other. Testing showed that data from the end nodes of the branch tunnel converges to the T-junction node after three hops, and then is integrated into the main line topology from the T-junction node, with no packet loss throughout the process. If the branch tunnel continues to extend, nodes can be added and the branch depth expanded in the same way, with no limit on the branch level.

[0096] IV. Personnel Location and Environmental Data Correlation

[0097] In this embodiment, the UWB positioning modules built into each monitoring node employ the Time Difference of Arrival (TDOA) positioning algorithm. Workers wear UWB positioning tags when entering the utility tunnel, and these tags transmit positioning pulses at a frequency of 1Hz. Each monitoring node receives the pulses and records the arrival timestamp, uploading the timestamp information to the central control system, which then calculates the tag's position. Field measurements show a positioning accuracy of 30 centimeters and a positioning update frequency of once per second. The system spatially correlates the real-time location coordinates of the workers with the gas concentration data collected by the nearest monitoring node. On the central control system's 3D display interface, the gas concentration distribution at various locations within the utility tunnel is visually displayed using color gradients, while simultaneously displaying the workers' positions and movement trajectories in real-time with dynamic icons. When the gas concentration in a certain area exceeds the limit or personnel enter a hazardous area, the system can simultaneously display the location of the hazard source and the location of the threatened personnel, providing complete situational information for emergency command.

[0098] This embodiment verifies the comprehensive technical effects of the present invention in terms of modular and flexible configuration, long-distance unlimited extension, multi-directional branch expansion, and integration of personnel positioning and environmental monitoring. The system ran continuously for 30 days, with a data collection integrity rate of 99.3% and a positioning success rate of 97.8%.

[0099] This invention is not limited to the specific embodiments described above. Any modifications made by those skilled in the art based on the above concept without creative effort are within the scope of protection of this invention.

Claims

1. An adaptive topology control method for monitoring nodes in confined spaces, characterized in that, Includes the following steps: Step 1: Conduct on-site survey of the target confined space to obtain information on the spatial morphology, bifurcation node location, and occlusion area distribution of the confined space. Based on the survey results, divide the confined space into several sub-regions and configure the corresponding initial topology type for each sub-region. Step 2: Place monitoring nodes according to the initial topology of each sub-region. After the nodes are placed, each monitoring node is finely adjusted in position through a virtual force adaptive adjustment process. The virtual force includes the virtual repulsive force between adjacent monitoring nodes and the virtual attractive force of the covered void area on the monitoring node. Each monitoring node moves gradually according to the direction of the virtual resultant force until the full space coverage reaches the preset requirement. Step 3: Start all monitoring nodes. Each monitoring node autonomously determines its own working mode based on the energy state information and distribution density information of its surrounding one-hop neighbor nodes, and completes local clustering based on the working mode to generate a global topology. Step four: Each monitoring node collects environmental parameter data within the confined space and transmits the environmental parameter data to the central control system through the global topology.

2. The method according to claim 1, characterized in that, In step two, the virtual repulsive force acts between any two adjacent monitoring nodes whose actual distance is less than the preset optimal spacing, driving the two monitoring nodes to move away from each other. The virtual attraction acts between the monitoring node and a spatial area not covered by the perception range of any node, driving the monitoring node to move closer to that area. The virtual resultant force is determined by the vector superposition of the virtual repulsive force and the virtual attractive force.

3. The method according to claim 1, characterized in that, In step two, the method for determining whether the full space coverage rate meets the preset requirements is as follows: The finite space is discretized into several grid cells. The number of grid cells covered by the sensing range of at least one monitoring node is counted. The ratio of the number of covered grid cells to the total number of grid cells is calculated. When the ratio is greater than or equal to a preset coverage threshold, the virtual force adaptive adjustment process is determined to be complete.

4. The method according to claim 1, characterized in that, In step three, each monitoring node autonomously determines its own operating mode based on the energy state information and distribution density information of its surrounding one-hop neighbor nodes. Specifically: Each monitoring node obtains its own remaining energy and the average remaining energy of its one-hop neighbor nodes, compares its own remaining energy with the average remaining energy, and determines its own energy level. Each monitoring node obtains the number of one-hop neighbor nodes and calculates the distribution density of its local area in combination with its own communication coverage area; For monitoring nodes at high energy levels and whose local distribution density is below a preset lower limit, the routing backbone working mode is selected; Monitoring nodes located at low energy levels or with a local distribution density exceeding a preset upper limit should select the terminal acquisition mode; other monitoring nodes should select the relay forwarding mode.

5. The method according to claim 1, characterized in that, After the global topology is generated, each monitoring node continuously monitors the signal quality of the communication link. When any node fails or the signal quality of any communication link is lower than the preset quality standard, a local topology reconstruction is triggered. The scope of the local topology reconstruction is limited to the one-hop neighbors of the failed node or abnormal link, and after reconstruction, it is re-integrated into the global topology structure.

6. The method according to claim 1, characterized in that, In step one, the initial topology type includes linear chain topology and branch topology; The linear chain topology is configured for sub-regions with a long and narrow spatial shape, and the branching topology is configured for sub-regions with branching nodes.

7. The method according to claim 1, characterized in that, In step one, each sub-region is configured with a corresponding coverage redundancy parameter. The coverage redundancy is used to characterize the degree of overlap of the sensing range of the monitoring nodes in the sub-region. For sub-regions containing high-risk key monitoring areas marked in the survey results, a first coverage redundancy is configured, and for sub-regions containing non-high-risk key monitoring areas, a second coverage redundancy is configured, with the first coverage redundancy being greater than the second coverage redundancy.

8. The method according to claim 1, characterized in that, In step three, each monitoring node exchanges topology information only with one-hop neighbor nodes. Each monitoring node independently decides its own transmission power level based on the number and location information of one-hop neighbor nodes, and selects the lowest transmission power level while maintaining connectivity with at least one neighbor node.

9. The method according to claim 1, characterized in that, The monitoring nodes are modular nodes, including independently replaceable sensor modules, communication modules, and power supply modules. The sensor modules are equipped with different types and accuracies of sensors according to monitoring requirements. For sub-areas with multi-directional branch extension requirements, branch relay nodes or three-way switching nodes are configured. The branch relay nodes have multiple communication interfaces and can establish links with neighboring nodes in three or more directions simultaneously, realizing multi-level branch expansion in space. For sub-areas with a narrow and long orientation that require long-distance coverage, the network extends step by step along the length direction through multi-hop communication between adjacent nodes. New nodes can automatically discover and access the existing topology, realizing the infinite extension of the monitoring network in the length direction.

10. The method according to claim 1, characterized in that, The monitoring node also integrates a positioning module to obtain its own or personnel / equipment carrying positioning tags in real time. The location information is transmitted to the central control system through a global topology structure to realize real-time positioning and trajectory tracking of workers in a confined space. The environmental parameter data collected by each monitoring node is associated and packaged with the corresponding location information to form monitoring data with spatial coordinates.