An underground space mapping and environmental monitoring system and method

Abstract

An underground space mapping and environmental monitoring system includes a data acquisition unit, an edge computing unit, a dynamic scheduling unit, and a simulation optimization unit. The data acquisition unit acquires various parameter data. The edge computing unit receives the various parameter data and predicts future tasks, node energy consumption, and network congestion risks based on historical data and real-time status. The dynamic scheduling unit selects sensor nodes based on future tasks, node energy consumption, and network congestion risks, aggregates the data collected by the sensor nodes, and executes a low-latency routing algorithm to dynamically adjust and obtain a scheduling strategy. The simulation optimization unit verifies the performance of the scheduling strategy and feeds back optimization parameters to the dynamic scheduling unit to adjust the scheduling strategy and obtain an optimized scheduling strategy. This invention combines a multi-objective optimization strategy with an energy monitoring and scheduling mechanism to ensure that the network can achieve low-energy, high-efficiency real-time data transmission and monitoring tasks in complex underground space environments.

Description

Technical Field

[0001] This invention relates to the field of wireless sensor network technology, and more specifically, to an underground space mapping and environmental monitoring system and method based on low-latency routing algorithms and energy optimization strategies. Background Technology

[0002] With the expansion of underground space development and utilization, underground space mapping and environmental monitoring are becoming increasingly important in urban construction, mining, and tunnel construction. Traditional methods rely on manual labor or mechanical equipment, resulting in high energy consumption, low efficiency, and high data transmission latency. The complex underground environment makes the deployment and management of wireless sensor networks (WSNs) difficult, especially in confined spaces. Achieving efficient and stable data acquisition and transmission while reducing energy consumption and latency remains a significant technological challenge.

[0003] Currently, data transmission latency is high, network path optimization is not optimized, congestion is difficult to predict, and fixed topology increases latency; energy consumption is uneven, static energy management does not consider node state changes, leading to premature failure of some nodes; network congestion and data loss occur, existing congestion control cannot predict and adjust in real time, and it is difficult to alleviate congestion in a timely manner; topology change adaptability is poor, static routing algorithms cannot cope with dynamic changes in node state and location; task coordination and management efficiency is low, multi-device collaborative task scheduling lacks flexibility, and resource waste is serious, among other problems.

[0004] Therefore, the existing technology has problems and needs further improvement and development. Summary of the Invention

[0005] (I) Purpose of the invention: In order to solve the problems existing in the prior art, the purpose of the present invention is to provide an underground space mapping and environmental monitoring system and method, which solves the key problems of low latency, energy optimization and network stability in space mapping and environmental monitoring systems.

[0006] (II) Technical Solution: In order to solve the above-mentioned technical problems, this technical solution provides an underground space mapping and environmental monitoring system, including a data acquisition unit, an edge computing unit, a dynamic scheduling unit, and a simulation optimization unit;

[0007] The data acquisition unit includes multiple sensor nodes for acquiring various parameter data;

[0008] The edge computing unit receives various parameter data and predicts future tasks, node energy consumption, and network congestion risks based on historical data and real-time status.

[0009] The dynamic scheduling unit selects sensor nodes based on future tasks, node energy consumption, and network congestion risks, aggregates the collected data from the sensor nodes, executes a low-latency routing algorithm, and makes dynamic adjustments to obtain a scheduling strategy.

[0010] The simulation optimization unit verifies the performance of the scheduling strategy and feeds back optimization parameters to the dynamic scheduling unit to adjust the scheduling strategy and obtain the optimized scheduling strategy.

[0011] Preferably, sensor nodes are deployed in underground space using unmanned mobile devices to collect various parameter data; the parameter data includes environmental parameter data, spatial image data, and node status parameters.

[0012] Preferably, the edge computing unit first predicts the energy, bandwidth and computing resource requirements of the task based on historical task data, node energy status and real-time network status;

[0013] ,

[0014] in, This represents the remaining energy at time t; Represents the initial energy of the node; This indicates the energy consumption rate of the current node; This represents the time step, or the frequency of node energy updates.

[0015] Finally, based on node load, traffic, and topology changes, the network congestion risk is predicted, and the congestion area and degree of congestion are output.

[0016] Preferably, the low-latency routing algorithm includes a clustering routing submodule, an opportunistic routing submodule, and a multi-objective optimization submodule; the clustering routing submodule divides the sensor nodes into multiple clusters, and selects the globally optimal cluster head node based on the node's remaining energy, distance from the base station, and load using a particle swarm optimization algorithm;

[0017] The opportunistic routing submodule monitors the remaining energy, load, and distance to the target node in real time, and dynamically selects the node with sufficient energy and low load as the next-hop transmission path.

[0018] The multi-objective optimization submodule combines the minimum spanning tree algorithm with communication distance and energy consumption as dual objectives to optimize the transmission path between cluster heads and minimize the total transmission cost.

[0019] Preferably, the dynamic scheduling unit further includes a distributed predictive congestion control module; firstly, it collects the node buffer queue status, remaining energy, and distance to the target node in real time, and calculates the node transmission gravity; secondly, it dynamically adjusts the data flow based on the transmission gravity, and selects nodes with a gravity less than the first gravity threshold as transmission paths; finally, it combines a hybrid CSMA / TDMA protocol, using CSMA to avoid channel conflicts when the node gravity is less than the first gravity threshold, and using TDMA to allocate fixed time slots for transmission when the node gravity is greater than or equal to the first gravity threshold.

[0020] Preferably, the dynamic scheduling unit further includes a prediction and matching module, specifically including pre-deploying cluster head selection, path planning and task priority strategies based on the prediction results of the edge computing unit;

[0021] Compare the deviation between the preset strategy and the actual execution. If the time delay deviation is less than or equal to the first time delay threshold and the energy consumption deviation is less than or equal to the first energy threshold, execute the preset strategy.

[0022] When the time delay deviation is greater than the first time delay threshold and / or the energy consumption deviation is greater than the first energy threshold, a new cluster head is selected and the transmission path is updated by rerunning the PSO algorithm based on the current node status, network load and task priority.

[0023] Preferably, the simulation optimization unit is based on NS-3 and MATLAB tools to verify the scheduling strategy with multiple indicators, including simulating the network load of multiple nodes, testing end-to-end transmission latency; monitoring node energy consumption rate; simulating high-traffic scenarios, and testing packet loss rate and throughput.

[0024] The verification results are fed back to the dynamic scheduling unit to optimize the particle velocity parameters of the PSO algorithm and the weight coefficients of MST.

[0025] Preferably, it also includes an energy management unit, which is connected to the data acquisition unit and the dynamic scheduling unit; including adjusting the sensor node acquisition frequency; controlling the node transmission power; triggering the sleep mode of low-energy nodes; activating the energy supply of high-priority tasks when the task priority decreases, and delaying the execution of low-priority tasks.

[0026] Preferably, the data acquisition unit, edge computing unit, and dynamic scheduling unit communicate with each other through a wireless sensor network, using a hybrid CSMA / TDMA protocol; when the node utilization rate is less than or equal to the first node utilization rate threshold, the CSMA mechanism is used, the node listens to the channel before transmission, and transmits immediately when idle;

[0027] When the node utilization rate is greater than the first node utilization rate threshold, the TDMA mechanism is used to allocate a fixed transmission time slot to each node.

[0028] When the second node utilization threshold is reached, the dynamic scheduling unit automatically triggers protocol switching based on the real-time node utilization; the second node utilization threshold is greater than the first node utilization threshold.

[0029] A method for underground space mapping and environmental monitoring, applicable to an underground space mapping and environmental monitoring system, specifically includes:

[0030] Step 1: Multiple sensor nodes acquire various parameter data;

[0031] Step 2: The edge computing unit receives various parameter data and predicts future tasks, node energy consumption, and network congestion risks based on historical data and real-time status.

[0032] Step 3: The dynamic scheduling unit selects sensor nodes based on future tasks, node energy consumption, and network congestion risks, aggregates the data collected by the sensor nodes, executes a low-latency routing algorithm, and makes dynamic adjustments to obtain the scheduling strategy.

[0033] Step 4: The simulation optimization unit verifies the performance of the scheduling strategy and feeds back the optimization parameters to the dynamic scheduling unit to adjust the scheduling strategy and obtain the optimized scheduling strategy.

[0034] (III) Beneficial Effects: This invention provides an underground space mapping and environmental monitoring system and method. First, by predicting tasks and network status, it pre-deploys task strategies, dynamically adjusts routes and task scheduling, significantly reducing data transmission latency and enabling the system to respond promptly and efficiently complete real-time tasks in complex and dynamic underground environments. Second, it introduces an intelligent energy management module to monitor node energy status and consumption rates in real time, and optimizes node operating frequency and data transmission strategies based on a prediction mechanism, extending node lifespan and preventing task failures or network interruptions due to energy depletion. Third, through distributed predictive congestion control (DPCC) and dynamic route adjustment, it effectively predicts potential congestion areas in the network, dynamically adjusts data traffic and path selection, and avoids the problems of inflexible path selection or insufficient congestion control in existing technologies. It can dynamically adjust routing strategies and task priorities according to changes in task requirements, node status, and network load, and flexibly respond to frequent changes in network topology in confined environments such as underground spaces through multi-objective optimization routing algorithms and prediction mechanisms. Finally, by using prediction and matching strategies, the optimal cluster head node and data transmission path are selected in advance according to task requirements, reducing redundant transmission, lowering latency and energy consumption, and avoiding task failure due to excessive node load or insufficient energy. Attached Figure Description

[0035] Figure 1 This is a schematic diagram of the structure of an underground space mapping and environmental monitoring system according to the present invention;

[0036] Figure 2This is a flowchart of the steps of an underground space mapping and environmental monitoring method according to the present invention. Detailed Implementation

[0037] The present invention will be further described in detail below with reference to preferred embodiments. More details are set forth in the following description in order to provide a full understanding of the present invention. However, the present invention can obviously be implemented in many other ways different from those described herein. Those skilled in the art can make similar extensions and derivations based on actual application situations without departing from the spirit of the present invention. Therefore, the scope of protection of the present invention should not be limited by the content of this specific embodiment.

[0038] The accompanying drawings are schematic diagrams of embodiments of the present invention. It should be noted that these drawings are for illustrative purposes only and are not drawn to scale, and should not be construed as limiting the actual scope of protection of the present invention.

[0039] An underground space mapping and environmental monitoring system, such as Figure 1 As shown, the system includes a data acquisition unit, an edge computing unit, a dynamic scheduling unit, a simulation optimization unit, and an energy management unit. The data acquisition unit includes multiple sensor nodes for acquiring various parameter data. The edge computing unit receives this parameter data and predicts future tasks, node energy consumption, and network congestion risks based on historical data and real-time status. The dynamic scheduling unit selects sensor nodes based on future tasks, node energy consumption, and network congestion risks, aggregates the collected data from these nodes, and executes a low-latency routing algorithm to dynamically adjust and obtain a scheduling strategy. The simulation optimization unit verifies the performance of the scheduling strategy and feeds back optimization parameters to the dynamic scheduling unit to adjust the scheduling strategy and obtain an optimized scheduling strategy. The energy management unit is connected to both the data acquisition unit and the dynamic scheduling unit; the energy management unit includes adjusting the sensor node acquisition frequency; controlling node transmission power; triggering a sleep mode for low-energy nodes; and activating a sleep mode for low-priority nodes when task priority decreases to prioritize energy supply for high-priority tasks, while delaying the execution of low-priority tasks.

[0040] A method for underground space mapping and environmental monitoring, applicable to an underground space mapping and environmental monitoring system, such as... Figure 2 As shown, the specific steps include:

[0041] Step 1: Multiple sensor nodes acquire various parameter data;

[0042] Step 2: The edge computing unit receives various parameter data and predicts future tasks, node energy consumption, and network congestion risks based on historical data and real-time status.

[0043] Step 3: The dynamic scheduling unit selects sensor nodes based on future tasks, node energy consumption, and network congestion risks, aggregates the data collected by the sensor nodes, executes a low-latency routing algorithm, and makes dynamic adjustments to obtain the scheduling strategy.

[0044] Step 4: The simulation optimization unit verifies the performance of the scheduling strategy and feeds back the optimization parameters to the dynamic scheduling unit to adjust the scheduling strategy and obtain the optimized scheduling strategy.

[0045] More specifically, the data acquisition unit includes an unmanned mobile device and various sensor nodes; the sensor nodes are deployed in underground space via the unmanned mobile device. The sensor nodes collect various parameter data, including environmental parameter data, spatial image data, and node status parameters.

[0046] The unmanned mobile devices include drones, unmanned vehicles, quadruped robots, etc. Each unmanned mobile device has autonomous navigation, positioning, and sensor data acquisition functions, enabling it to flexibly cope with complex environments. The sensor nodes include sensors such as temperature and humidity sensors, gas concentration sensors, and pressure sensors, as well as imaging devices such as lidar and cameras.

[0047] A sensor network is composed of multiple sensor nodes, each with wireless communication capabilities. Each sensor node includes a sensor and an imaging device; the sensors collect underground environmental data and monitor node status in real time, while the imaging devices acquire underground space data in real time, ensuring comprehensive and accurate acquisition of information about the underground space.

[0048] More specifically, the edge computing unit's specific operations include:

[0049] The edge computing unit includes edge computing nodes set around the sensor nodes. After receiving various parameter data, the edge computing nodes perform data preprocessing and feature extraction, and based on task prediction and network status analysis, they also predict future task requirements, node energy status, network congestion, etc.

[0050] Step 201: Perform data preprocessing on the parameter data through edge computing nodes to obtain unified parameter data.

[0051] Specifically, the system receives and collects parameter data, then performs data preprocessing, including data cleaning, data formatting, and data aggregation. Data cleaning removes noisy data and outliers, and can be achieved by using filtering algorithms (such as Kalman filtering) to remove invalid data caused by transmission interference and equipment errors.

[0052] The data formatting process transforms the acquired parameter data into a unified format according to different task requirements, facilitating subsequent analysis and calculations. Data standardization eliminates dimensional differences (e.g., unifying the units of temperature, humidity, and pressure data) to ensure the accuracy of subsequent analysis.

[0053] The data aggregation performs spatiotemporal aggregation on parameter data of the same type (such as fusion of data from multiple nodes in the same region), reducing redundant transmission. The data aggregation may employ a lightweight compression algorithm to compress the preprocessed data, reducing transmission bandwidth requirements.

[0054] Real-time preprocessing via the edge computing nodes can reduce latency.

[0055] Step 202: Perform data fusion and feature extraction on the unified parameter data to optimize the resource requirements and path of the computation task.

[0056] Based on historical task data, node energy consumption, network status, and other information, predict future task resource requirements, latency requirements, priorities, etc. Use machine learning algorithms and historical data analysis methods to generate predictive strategies for task execution, including task scheduling and data transmission paths.

[0057] Specifically, the lidar point cloud data is voxelized to convert the 3D spatial data into a regular voxel mesh, and spatial geometric features (such as obstacle contours and spatial topology) are extracted. Edge detection and target recognition (such as using the HOG+SVM algorithm) are performed on the camera image data to extract object features in the environment (such as pipes, cracks, and structural contours).

[0058] Extract the state parameters (remaining energy, communication signal strength, load rate) of sensor nodes to form a node health status feature vector. Analyze network topology features (such as node connectivity and cluster head node load distribution) to identify potential network bottlenecks.

[0059] Transforming raw data into a more informative and easily transmittable format reduces data redundancy and optimizes data flow. Feature extraction can extract key features such as the shape, position, and size of objects from the data, while voxelization further transforms spatial data into voxel format, improving data compression ratio and processing efficiency.

[0060] Step 203: Based on the resource requirements and path optimization of the task, predict the future energy consumption of the task and nodes, as well as the risk of network congestion.

[0061] Specifically, edge computing nodes, based on task prediction and network status analysis, and combining historical and real-time data, predict task resource requirements, execution time periods, network load, etc., generate task scheduling and cluster head selection schemes, and preset data transmission paths.

[0062] Based on historical mission data and real-time environmental data, the resource requirements of the mission (such as energy, bandwidth, computing power, etc.) are predicted, as shown in formula (1):

[0063] (1),

[0064] in, Indicates the task's resource requirements (including energy, bandwidth, etc.); Represents historical task data (such as execution time, resource consumption, and latency); This indicates the node's current energy consumption rate; It represents real-time information such as node load and network status; Δt represents the time window, the expected time for task execution.

[0065] The node energy state is shown in formula (2):

[0066] (2),

[0067] in, This represents the remaining energy at time t; Represents the initial energy of the node; This indicates the energy consumption rate of the current node; This represents the time step, or the frequency of node energy updates.

[0068] Finally, based on node load, traffic, and topology changes, the network congestion risk is predicted, and the congestion area and degree of congestion are output, as shown in formula (3):

[0069] (3),

[0070] in, This indicates network load (predicted congestion). This indicates the load status of the node; This represents the node's traffic (the amount of data transmitted). Indicates changes in network topology (node ​​connectivity, signal strength, etc.); This indicates the time window, representing the predicted duration of network load.

[0071] More specifically, the dynamic scheduling unit's specific operations include:

[0072] Based on predicted task requirements, network status, node energy, and other information, the system selects appropriate cluster head nodes and data transmission paths, and dynamically adjusts task priorities and execution order. This ensures tasks are executed smoothly according to priority, avoiding delays or failures caused by rapid energy consumption or network congestion. The system dynamically selects cluster heads and deploys appropriate path selection strategies based on task priority, node energy status, and network load. The execution order and paths of tasks are adjusted in real time to ensure timely task completion and minimize data transmission latency.

[0073] The dynamic scheduling unit includes a low-latency routing module, a distributed predictive congestion control module, and a prediction and matching scheduling module. The low-latency routing module includes a clustering routing submodule, an opportunistic routing submodule, and a multi-objective optimization submodule. Based on the remaining energy of nodes, transmission distance, and network load, it dynamically selects cluster head nodes and data transmission paths by combining clustering routing, opportunistic routing, particle swarm optimization (PSO), and minimum spanning tree (MST) algorithms to minimize data transmission latency.

[0074] The clustering routing submodule divides sensor nodes into multiple clusters. In the initial stage of routing, sensor nodes are grouped, and each group selects a cluster head node. The cluster head node is responsible for data aggregation and transmission. Through the clustering routing algorithm, redundant data transmission and uneven load distribution can be reduced, and network energy efficiency can be improved.

[0075] By grouping sensor nodes into clusters and selecting cluster head nodes for data aggregation and forwarding, redundant transmissions in the network are reduced, improving energy efficiency. In energy-optimized low-latency routing algorithms, clustering algorithms can effectively reduce redundant communication between nodes and reduce energy consumption by aggregating data, thereby extending network lifetime.

[0076] Sensor nodes are deployed and grouped according to certain rules to form multiple clusters. Each cluster is managed by a cluster head node. The cluster head selection process is based on the node's remaining energy and distance from the base station, ensuring that the cluster head is energy-optimized while preventing some cluster heads from prematurely exhausting their energy due to overload. Each cluster head collects and aggregates data from all sensor nodes in its cluster, reducing redundant data transmission. The aggregated data is then transmitted to the base station via an optimized transmission path.

[0077] The energy monitoring module monitors the entire network's energy consumption in real time, ensuring that each node and cluster head completes its tasks before running out of power. The energy optimization module adjusts data acquisition frequency and transmission strategies to reduce energy consumption and extend network lifespan. Based on each node's energy and distance information, it dynamically selects transmission paths. Path optimization considers both energy and distance—two key factors—to ensure reduced latency and energy consumption during data transmission.

[0078] In this invention, the core objective of the clustering routing algorithm is to improve the performance of wireless sensor networks in confined spaces (such as mines and tunnels) by optimizing energy consumption and latency. The main steps of the algorithm are as follows:

[0079] Sensor nodes are dynamically grouped within the network, with each group managed by a cluster head node. Using a clustering algorithm, nodes are assigned to different clusters based on their geographical location, energy status, and communication load. This process effectively reduces redundant data transmission in the network and distributes the load evenly across the cluster head nodes, preventing some cluster head nodes from running out of energy due to overload.

[0080] The selection of cluster head nodes is based on factors such as the node's remaining energy, distance from the base station, and network load. The algorithm prioritizes nodes with sufficient energy and close proximity to the base station as cluster heads to ensure efficient data transmission and low latency. The cluster head's role is not only to aggregate data but also to transmit it; therefore, its energy consumption and load management are crucial.

[0081] The cluster head node is responsible for receiving and aggregating data from sensor nodes within its cluster. Data aggregation helps reduce redundant transmissions and optimizes network bandwidth usage. After aggregating the data, the cluster head transmits it to the base station via a selected optimal path. The selection of the transmission path is based on factors such as the distance between the node and the base station, the cluster head's remaining energy, and network load.

[0082] To further reduce transmission latency and energy consumption, the algorithm optimizes data transmission by dynamically selecting the best transmission path. Path optimization considers not only the energy state of nodes but also the distance between each node and the cluster head to adjust the data transmission path. Nodes with sufficient energy and low communication load are prioritized as relay nodes to avoid network bottlenecks or node overload.

[0083] The energy monitoring module monitors the energy consumption of each node and cluster head in the network in real time. When the energy of a node or cluster head drops to a preset threshold, the algorithm automatically adjusts the node's operating status, such as reducing data acquisition frequency and transmission power, thereby optimizing energy use and extending network lifespan.

[0084] To prevent some cluster heads from depleting their energy too quickly and impacting network performance, the algorithm also incorporates a node rotation mechanism. By periodically updating cluster head roles, it ensures that energy consumption across all nodes is as balanced as possible, preventing network outages or transmission failures.

[0085] Clustering routing algorithms combine multiple strategies, including node grouping, cluster head selection, data aggregation, path optimization, and energy management, to achieve low latency and energy optimization in confined spaces. This algorithm not only effectively reduces network congestion but also maximizes network operating efficiency with limited energy resources.

[0086] The opportunistic routing submodule monitors the remaining energy, load, and distance to the target node in real time, and dynamically selects the node with sufficient energy and low load as the next-hop transmission path.

[0087] Opportunistic routing selects the next-hop node based on the node's remaining energy and distance to the sink node, transmitting data at appropriate times to balance network energy consumption. In energy-optimized low-latency routing algorithms, the combination of opportunistic routing and clustering algorithms can balance energy allocation, reduce node load, thereby lowering latency and extending network lifespan.

[0088] Each sensor node monitors its remaining energy, distance to the cluster head node, and current load in real time. This information forms the basis of the opportunistic routing algorithm, determining whether to select that node as part of the data transmission path. Based on node status monitoring, the routing module calculates and selects the optimal transmission path for each node, considering factors such as its remaining energy, load, and distance to the target node. The algorithm dynamically adjusts its routing strategy to ensure minimal network energy consumption and latency. Once the path selection is complete, the opportunistic routing algorithm chooses the optimal next-hop node for data transmission based on the current network state and the selected path. In this way, data transmission can occur between nodes with sufficient energy and low load, thus avoiding network congestion and energy depletion.

[0089] The network's energy usage is tracked in real time by the energy monitoring module, which adjusts the data acquisition frequency and transmission strategy based on the energy consumption of each node. The energy optimization module ensures that each node completes its tasks before its battery runs out, avoiding network outages caused by node energy depletion.

[0090] The core objective of opportunistic routing algorithms is to achieve low-latency, energy-optimized data transmission in wireless sensor networks within confined spaces by dynamically adjusting routing paths. The algorithm is designed to select the optimal transmission path based on the real-time status of sensor nodes (such as remaining energy, communication load, and distance to the target node), ensuring data transmission in the shortest possible time while minimizing energy consumption.

[0091] Each sensor node periodically monitors its remaining energy, communication load, and distance to other nodes. A node's remaining energy is a key factor in the algorithm's decision-making; nodes with sufficient energy are prioritized for data transmission. This information is transmitted to the cluster head node or the network optimization module for subsequent routing decisions.

[0092] Based on the state information of each node, the algorithm dynamically selects the next-hop node. Routing selection considers not only the node's remaining energy but also its communication load and distance. The algorithm prioritizes nodes with sufficient remaining energy and close proximity to the target node as data relay nodes to avoid energy depletion or network congestion. After selecting the optimal next-hop node, data is transmitted through the chosen path. The transmission path is dynamically adjusted based on the real-time network status. Unlike static routing algorithms, opportunistic routing algorithms can adapt to changes in network conditions, ensuring data transmission is completed efficiently and with low energy consumption.

[0093] The energy monitoring module tracks the energy consumption of each node in real time and adjusts the data acquisition frequency and transmission strategy based on the monitoring results. For nodes that consume energy too quickly, the system will reduce their activity frequency or delay data transmission to prevent node failure or network interruption due to energy depletion.

[0094] Opportunistic routing algorithms employ a feedback mechanism, optimizing and adjusting the routing strategy after each transmission based on the current network state (such as network congestion and changes in node energy). Through simulation and real-time feedback, the algorithm can continuously improve data transmission paths, ensuring continuous optimization of overall network performance.

[0095] Opportunistic routing algorithms, through real-time monitoring of node status, intelligent route selection, dynamic data transmission path adjustment, and energy optimization, can achieve efficient, low-latency data transmission in complex, confined spaces, while extending network lifespan. This algorithm is particularly suitable for wireless sensor network applications requiring low latency and high energy efficiency, such as underground space mapping and environmental monitoring.

[0096] The multi-objective optimization submodule combines particle swarm optimization (PSO) and minimum spanning tree (MST) algorithms to optimize the transmission path between cluster heads and minimize the total transmission cost, with communication distance and energy consumption as dual objectives.

[0097] By simulating particle swarm optimization (PSO) behavior, cluster head selection is optimized, taking into account factors such as node remaining energy and distance from the base station to ensure globally optimal cluster head selection. In the improved LEACH algorithm, the PSO algorithm optimizes cluster head selection, avoids uneven energy consumption, and improves network energy efficiency.

[0098] In particle swarm optimization (PSO), the position and velocity of the particles must first be initialized. Each particle represents a candidate solution (i.e., cluster head selection or path selection), and its position and velocity reflect the current state of the solution. The initial position is usually randomly selected, and the velocity is adjusted as needed to find the optimal solution in the search space.

[0099] For each particle, the quality of its solution is evaluated by calculating its fitness. The fitness function evaluates the particle's performance based on factors such as the energy efficiency of cluster head selection, transmission path latency, and network load. Particles with higher fitness represent better solutions (i.e., cluster heads or paths with high energy efficiency and low latency).

[0100] Based on each particle's fitness, its velocity and position are updated. The update rules are typically based on classic formulas from particle swarm optimization algorithms, where the particle's position and velocity are influenced by both the current best position and the global best position. By continuously adjusting the particle's direction and velocity, the particle gradually approaches the optimal solution.

[0101] In particle swarm optimization, the particle with the highest fitness value is selected as the current optimal solution. In the cluster head selection problem, a particle with high fitness represents a cluster head selection scheme with the minimum energy consumption and shortest transmission delay; in the path selection problem, a particle with high fitness represents the optimal transmission path.

[0102] Throughout the process, the energy monitoring module tracks the energy consumption of each node and cluster head in real time, ensuring the system operates efficiently under energy constraints. The data flow optimization module optimizes data flow and transmission strategies based on the optimal path and cluster head selection scheme provided by the particle swarm optimization algorithm, thereby reducing energy consumption and improving transmission efficiency.

[0103] By employing the PSO algorithm, this invention enables efficient selection of cluster heads and transmission paths in wireless sensor networks, balancing energy consumption and latency. The Particle Swarm Optimization algorithm adaptively searches for the optimal solution and improves the network's energy efficiency and performance through continuous iteration. This design is suitable for various complex network environments, especially for wireless sensor network applications requiring efficient and stable communication in confined spaces.

[0104] By constructing a minimum spanning tree, the data transmission path between cluster heads is optimized, reducing communication distance and energy consumption. In the PSO-ACO-LEACH algorithm, the minimum spanning tree is used to select the optimal transmission path between cluster heads, thereby maximizing network energy efficiency.

[0105] In the Minimum Spanning Tree (MST) algorithm, the communication distance between cluster head nodes is calculated first. Each cluster head node calculates its communication cost based on its distance from other cluster head nodes. This distance is typically proportional to energy consumption, making the calculation of communication distance a crucial step in optimizing network performance.

[0106] Based on the communication distance information between cluster heads, the minimum spanning tree algorithm selects the paths with the lowest cost to construct the optimal transmission paths between cluster heads. The minimum spanning tree ensures that each cluster head node in the network can connect to other cluster head nodes via the shortest path, thereby reducing communication distance and energy consumption. After constructing the minimum spanning tree, the path optimization module further optimizes the transmission paths. The goal of path optimization is to reduce energy consumption by decreasing the communication distance between cluster heads, while ensuring the stability and efficiency of data transmission. The optimized paths not only improve network energy efficiency but also reduce data transmission latency.

[0107] The energy monitoring module tracks the energy consumption of each node and cluster head in real time. By adjusting the data acquisition frequency and transmission strategy, it ensures that cluster heads and sensor nodes operate stably for extended periods with limited energy resources. The data flow optimization module further optimizes data traffic and transmission strategies based on the optimized transmission path and cluster head selection strategy, ensuring improved overall network performance.

[0108] The Minimum Spanning Tree (MST) algorithm, as shown in Equation (4), aims to select an edge set E such that the total weight of the edges is such that... The minimum value is achieved while ensuring that all nodes in the graph are connected and do not form cycles.

[0109] (4),

[0110] Where MST represents the set of edges of the minimum spanning tree; This represents the set of all edges in the graph, where each edge connects two nodes. And has a weight ; Representing an edge The weight of the signal is usually represented by the communication distance or energy consumption.

[0111] By using the minimum spanning tree algorithm, this invention effectively reduces the communication distance between cluster heads, thereby minimizing energy consumption and transmission latency. This algorithm not only optimizes data transmission paths but also improves overall network performance, making it particularly suitable for energy-constrained and latency-sensitive wireless sensor network applications.

[0112] The distributed predictive congestion control module monitors network traffic, node load, and latency in real time to predict potential congestion areas and dynamically adjusts data traffic and transmission paths to avoid increased latency and data loss caused by congestion. It can predict and mitigate network congestion in advance, reducing latency and data loss.

[0113] Specifically, firstly, the node buffer queue status, remaining energy, and distance to the target node are collected in real time to calculate the node transmission gravity; secondly, the data flow is dynamically adjusted based on the transmission gravity, and nodes with a gravity less than the first gravity threshold are selected as transmission paths; finally, a hybrid CSMA / TDMA protocol is used, CSMA is used to avoid channel collisions when the node gravity is less than the first gravity threshold, and TDMA is used to allocate fixed time slots for transmission when the node gravity is greater than or equal to the first gravity threshold.

[0114] By predicting network congestion and dynamically adjusting it, the distribution of data traffic is controlled to avoid increased latency caused by network congestion. In low-latency routing algorithms based on congestion control, the DPCC mechanism is used to predict congestion, combined with a hybrid CSMA / TDMA mechanism to optimize data flow in the network and reduce latency. Congestion prediction: The first key step of DPCC is to predict potential future network congestion through the analysis of historical data. Each sensor node and cluster head node periodically reports network status to the control center, including information such as traffic, load, and latency. Based on this information, the DPCC module can assess potential congestion in the network and predict which areas or nodes may experience congestion. Based on the output of the congestion prediction module, the traffic adjustment module avoids network congestion by dynamically adjusting data traffic. This module adjusts the data transmission rate according to the real-time network status to avoid congestion caused by excessive data volume. By adjusting transmission time and node transmission frequency, network traffic is optimized to ensure timely data transmission.

[0115] When congestion occurs, the system selects an uncongested path for data transmission. By choosing less-loaded relay nodes or cluster heads, congested paths are avoided, thereby reducing latency and packet loss rate, ensuring data can be quickly transmitted to the target node. The DPCC system also works in conjunction with the energy management system to monitor the energy consumption of nodes and cluster heads in real time. When node energy is low, its data acquisition frequency and transmission strategy are adjusted to avoid excessive transmission on high-energy-consuming nodes, ensuring long-term stable network operation. The data flow optimization module dynamically adjusts data flow and transmission paths based on node energy status and network traffic to balance energy consumption and data transmission efficiency.

[0116] Predict potential future congestion based on historical traffic and load data. Dynamically adjust data traffic according to the predicted congestion information to avoid excessive network load. Select uncongested paths for data transmission to ensure smooth data transfer.

[0117] By combining energy monitoring, the working status of nodes and cluster heads can be optimized, thereby extending the network's lifespan.

[0118] DPCC can prevent network congestion and reduce latency and packet loss by predicting network congestion in advance. Simultaneously, by dynamically adjusting data flow and transmission paths, DPCC effectively optimizes network energy consumption, ensuring efficient and stable network operation in confined environments. This algorithm is particularly suitable for complex and dynamically changing network environments, such as wireless sensor networks in confined spaces like mines and tunnels.

[0119] The prediction and matching module pre-deploys task strategies (including cluster head selection, path planning, and task priority) based on the prediction results of the edge computing processing module. If the deviation between the prediction and the actual situation is within the tolerance range, the preset strategy is executed; otherwise, the optimal strategy is recalculated based on the current state. This strategy performs task prediction and dynamic adjustment based on historical task data and real-time environmental information, achieving more efficient task scheduling and path selection, and significantly improving the system's real-time performance, energy efficiency, and stability.

[0120] This includes pre-deploying cluster head selection, path planning, and task priority strategies based on prediction results from edge computing units; comparing the deviation between the preset strategy and the actual execution; if the latency deviation is less than or equal to a first latency threshold and the energy consumption deviation is less than or equal to a first energy threshold, executing the preset strategy; if the latency deviation is greater than the first latency threshold and / or the energy consumption deviation is greater than the first energy threshold, based on the current node status, network load, and task priority, selecting a new cluster head by re-running the PSO algorithm and updating the transmission path.

[0121] The most suitable cluster head node is selected based on factors such as the node's remaining energy, distance from the base station, and node load, as shown in formula (5):

[0122] (5),

[0123] in, Indicates the optimal cluster head selection; This represents the remaining energy of node i; This represents the distance between node i and the base station; This indicates the load status of node i.

[0124] As shown in formula (6), the optimal data transmission path is selected based on the node's remaining energy, load status, and distance:

[0125] (6),

[0126] in, This indicates the optimal path selected; This represents the distance (transmission path length) to node j. This represents the remaining energy of node j.

[0127] As shown in formula (7), the priority of tasks is dynamically adjusted based on the urgency of the task, the energy consumption rate of the node, the network load, etc.

[0128] (7),

[0129] in, Indicates the priority of the task; Indicates the urgency of the task; Represents node energy; Indicates network load; This represents the weighting coefficient, which is adjusted according to different environments and task types.

[0130] The core objective is to make decisions and deployments in advance within the system by predicting task information, network conditions, and node status, thereby achieving efficient task scheduling, path selection, and low-latency data transmission. This strategy leverages the real-time computing capabilities of edge computing nodes to predict task execution based on historical data and real-time status. Based on the prediction results, it selects appropriate routing strategies and task arrangements to ensure the system can respond in real-time in confined space environments, reducing energy consumption and improving data transmission efficiency.

[0131] The energy management unit is connected to the data acquisition unit and the dynamic scheduling unit; it includes adjusting the sensor node acquisition frequency; controlling the node transmission power; triggering the sleep mode of low-energy nodes; activating the energy management unit when the task priority decreases to prioritize the energy supply of high-priority tasks, while delaying the execution of low-priority tasks. The energy management unit monitors the energy consumption rate of each node in real time and dynamically adjusts the data acquisition frequency, transmission power, and operating mode to balance node energy consumption and extend the network lifetime.

[0132] The data acquisition unit, edge computing unit, and dynamic scheduling unit communicate via a wireless sensor network, employing a hybrid CSMA / TDMA protocol. When node utilization is less than or equal to a first node utilization threshold, the CSMA mechanism is used, with nodes listening to the channel before transmission and transmitting immediately when idle. When node utilization is greater than the first node utilization threshold, the TDMA mechanism is used, allocating a fixed transmission time slot to each node. When a second node utilization threshold is reached, the dynamic scheduling unit automatically triggers protocol switching based on real-time node utilization; the second node utilization threshold is greater than the first node utilization threshold.

[0133] The simulation optimization module uses tools such as NS-3 and MATLAB to simulate and verify low-latency routing algorithms, energy optimization strategies, and congestion control mechanisms, and feeds back optimization parameters to the dynamic routing scheduling module and energy management module to improve the system's adaptability in complex underground environments.

[0134] The simulation optimization unit, based on NS-3 and MATLAB tools, verifies the scheduling strategy using multiple metrics, including simulating the network load of multiple nodes, testing end-to-end transmission latency, monitoring node energy consumption rates, simulating high-traffic scenarios, and testing packet loss rate and throughput. The verification results are fed back to the dynamic scheduling unit to optimize the particle velocity parameters of the PSO algorithm and the weight coefficients of the MST.

[0135] As shown in formula (8), the task scheduling is optimized by adjusting the feedback based on the real-time status:

[0136] (8),

[0137] in, This indicates the adjusted resource allocation and task strategy; Indicates the current resource allocation and task status; Indicates the change in transmission delay; Indicates changes in node load; This indicates the change in the remaining energy of a node.

[0138] Example 1: Applied to underground space mapping system.

[0139] The underground space mapping system rapidly deploys wireless sensor nodes using unmanned equipment (such as drones, unmanned vehicles, and quadruped robots), while simultaneously collecting geographic location data of underground spaces using devices such as lidar and cameras. This collected data undergoes preliminary processing on edge devices (such as unmanned equipment or edge computing nodes) and is then transmitted back to the control center via a wireless sensor network (WSN). Unmanned equipment rapidly deploys sensor nodes in confined environments using the WSN to collect data on environmental conditions such as temperature, humidity, gas concentration, and pressure, laying the foundation for subsequent environmental monitoring and spatial analysis. The collected raw data is not only transmitted to edge computing nodes for processing but also undergoes feature extraction and voxelization, transforming the raw data into a more informative and easily transmitted format, reducing data redundancy and optimizing data flow. The feature extraction stage extracts key features such as the shape, location, and size of objects from the data, while voxelization further transforms the spatial data into voxel format, improving data compression ratio and processing efficiency.

[0140] After these processes, the data is optimized by a low-latency routing algorithm and quickly and efficiently transmitted back to the control center. During this process, the low-latency routing algorithm ensures data transmission is completed in the shortest possible time. Simultaneously, in conjunction with the energy management module, the data acquisition frequency and transmission strategy are dynamically adjusted to reduce equipment energy consumption and maximize its operating time. The energy management module also monitors the equipment's energy consumption in real time, ensuring system stability during extended operation and preventing interruptions due to rapid battery depletion.

[0141] With the support of simulation and optimization systems (such as NS-3 and MATLAB), this invention can be optimized before actual operation, ensuring that the system can adjust tasks and data transmission strategies according to actual needs under different environmental conditions, thereby maximizing system efficiency and transmission quality. Finally, the control center, through data analysis and decision-making, monitors the status of the underground space in real time, ensuring that the system can perform spatial mapping and real-time environmental monitoring efficiently and at low cost in complex and constrained spatial environments.

[0142] This system integrates edge computing, low-latency routing algorithms, energy management, and data optimization technologies to enable efficient collaborative operation of unmanned equipment in underground spaces. This makes underground space mapping and environmental monitoring more accurate, flexible, and scalable, greatly improving the efficiency and quality of mapping work. It is especially suitable for complex underground environments that are difficult to reach using traditional methods, such as mines, tunnels, and wartime bunkers.

[0143] This system's low-latency routing algorithm optimizes data transmission paths in wireless sensor networks by combining clustering routing, opportunistic routing, and load balancing strategies, aiming to minimize latency and reduce energy consumption. First, the clustering routing algorithm divides sensor nodes into multiple clusters and selects appropriate cluster head nodes based on the nodes' remaining energy, distance from the base station, and communication load. The cluster head nodes are responsible for data aggregation and transmission, thereby reducing redundant data transmission and effectively lowering energy consumption and latency.

[0144] Secondly, the opportunistic routing algorithm dynamically selects the data transmission path based on the real-time status of each node in the network (such as remaining energy, load, and distance from the target node). This algorithm ensures that nodes with sufficient energy and low network load are used as next-hop nodes as much as possible during data transmission, thereby avoiding transmission failures or increased latency caused by some nodes running out of energy or becoming congested.

[0145] To further improve system performance, the load balancing algorithm avoids overload on a single cluster head node by evenly distributing data streams. This balances the load across the entire network, reducing network bottlenecks and ensuring stable data transmission with low latency.

[0146] Based on this, the energy management module monitors the energy consumption of each node in real time and dynamically adjusts the data acquisition frequency and transmission strategy of the nodes to ensure that each device can continue to operate under limited energy resources. The data backhaul mechanism transmits data from the sensor nodes to the base station through an optimized low-latency routing algorithm, ensuring that the data can be transmitted back to the control center in the shortest possible time.

[0147] By combining simulation systems (such as NS-3 and MATLAB), the entire low-latency routing algorithm can be verified and optimized in different network environments, ensuring optimal performance in various application scenarios, meeting both real-time requirements and ensuring efficient energy use. This design is particularly suitable for sensor network applications with strict requirements for latency and energy efficiency, such as underground space mapping, smart city monitoring, and environmental monitoring. The underground space mapping system of this invention adopts a multi-objective optimized low-latency routing algorithm, combining low latency, energy optimization, and congestion control strategies to achieve efficient and stable underground space data transmission. The system rapidly deploys wireless sensor nodes using unmanned equipment (such as drones, unmanned vehicles, quadruped robots, etc.) and collects geographical location data of underground space using devices such as LiDAR and cameras. This data is first pre-processed on edge computing nodes, including feature extraction and voxelization, to reduce data volume and increase information density. Then, the processed data is transmitted through an optimized low-latency routing algorithm, which comprehensively considers latency, energy consumption, and congestion in the network, dynamically selecting the optimal transmission path to ensure data transmission in the shortest possible time and minimize energy consumption.

[0148] This routing algorithm employs a multi-objective optimization mechanism, balancing energy consumption and network load while ensuring low-latency data transmission. By introducing a congestion control strategy, the algorithm can predict and adjust data flow based on real-time network conditions, avoiding delays caused by network congestion. Data collected by sensor nodes is transmitted back to the base station via a wireless sensor network (WSN), and the optimized routing path ensures fast and efficient data transmission. Simultaneously, the energy management module monitors the device's energy consumption in real time, adjusting data acquisition frequency and transmission strategies to ensure energy efficiency during long-term operation and extend the system's lifespan.

[0149] With the support of simulation and optimization systems (such as NS-3 and MATLAB), this system can perform performance verification before practical application, ensuring efficient operation under different environmental conditions. Finally, the control center is responsible for receiving and analyzing data, and adjusting the mapping tasks based on real-time feedback to ensure the stability and efficiency of the entire system. This system provides a low-cost, high-efficiency solution for underground space mapping and real-time environmental monitoring, particularly suitable for complex and confined environments such as mines and tunnels. The multi-objective optimization low-latency routing algorithm of this invention achieves efficient data transmission and network optimization in underground space mapping systems by combining low-latency, energy optimization, and congestion control strategies. The system utilizes unmanned equipment (such as drones, unmanned vehicles, and quadruped robots) to rapidly deploy sensor nodes and collect geographical location data of underground space through devices such as LiDAR and cameras. The data collected by these sensor nodes is first transmitted to the cluster head node for aggregation processing, and then optimized through the low-latency routing algorithm to select the optimal transmission path, transmitting data to the base station with minimal latency. The low-latency routing algorithm combines energy optimization and congestion control strategies to ensure maximum energy utilization efficiency while reducing latency and avoiding network congestion.

[0150] During algorithm implementation, the energy optimization module reduces unnecessary energy consumption and extends equipment life by adjusting data acquisition frequency and transmission strategies. The congestion control module monitors network status in real time and predicts and adjusts data flow to avoid delays and packet loss caused by congestion. Through these optimization mechanisms, the system maintains efficient data transmission and low-energy operation, ensuring smooth mapping tasks, especially in complex and constrained underground environments.

[0151] This invention discloses an underground space mapping and environmental monitoring system and method. First, by predicting tasks and network status, it pre-deploys task strategies, dynamically adjusts routing and task scheduling, significantly reducing data transmission latency and enabling the system to respond promptly and efficiently complete real-time tasks in complex and dynamic underground environments. Second, it introduces an intelligent energy management module to monitor node energy status and consumption rates in real time. Combined with a predictive mechanism, it optimizes node operating frequency and data transmission strategies, extending node lifespan and preventing task failures or network interruptions due to energy depletion. Third, through distributed predictive congestion control (DPCC) and dynamic routing adjustment, it effectively predicts potential congestion areas in the network, dynamically adjusting data traffic and path selection, avoiding the problems of inflexible path selection or insufficient congestion control in existing technologies. It can dynamically adjust routing strategies and task priorities based on task requirements, node status, and network load, flexibly responding to frequent changes in network topology in confined environments such as underground spaces through multi-objective optimization routing algorithms and predictive mechanisms. Finally, through predictive and matching strategies, it pre-selects the optimal cluster head node and data transmission path based on task requirements, reducing redundant transmission, lowering latency and energy consumption, and preventing task failures due to excessive node load or insufficient energy.

[0152] The above description illustrates preferred embodiments of the present invention and helps those skilled in the art to more fully understand the technical solution of the present invention. However, these embodiments are merely illustrative and should not be construed as limiting the specific implementation of the present invention to these embodiments. For those skilled in the art, several simple deductions and modifications can be made without departing from the inventive concept, and all such modifications should be considered within the protection scope of the present invention.

Claims

1. An underground space mapping and environmental monitoring system, characterized in that, It includes a data acquisition unit, an edge computing unit, a dynamic scheduling unit, and a simulation optimization unit; The data acquisition unit includes multiple sensor nodes for acquiring various parameter data; The edge computing unit receives various parameter data and predicts future tasks, node energy consumption, and network congestion risks based on historical data and real-time status. It also uses machine learning algorithms and historical data analysis methods to generate predictive strategies for task execution. The dynamic scheduling unit selects sensor nodes based on future tasks, node energy consumption, and network congestion risks, aggregates the collected data from the sensor nodes, and executes a low-latency routing algorithm to dynamically adjust and obtain a scheduling strategy. The low-latency routing algorithm includes a clustering routing submodule, an opportunistic routing submodule, and a multi-objective optimization submodule. The clustering routing submodule divides the sensor nodes into multiple clusters and selects the globally optimal cluster head node based on the node's remaining energy, distance from the base station, and load using a particle swarm optimization algorithm. The opportunistic routing submodule monitors the remaining energy, load, and distance to the target node in real time, and dynamically selects the node with sufficient energy and low load as the next-hop transmission path. The multi-objective optimization submodule combines the minimum spanning tree algorithm with communication distance and energy consumption as dual objectives to optimize the transmission path between cluster heads and minimize the total transmission cost. The simulation optimization unit verifies the performance of the scheduling strategy based on NS-3 and MATLAB tools. The verification results are fed back to the dynamic scheduling unit to optimize the particle velocity parameters of the PSO algorithm and the weight coefficients of MST. The data acquisition unit, edge computing unit, and dynamic scheduling unit communicate with each other through a wireless sensor network, using a hybrid CSMA / TDMA protocol. When the node utilization rate is less than or equal to the first node utilization rate threshold, the CSMA mechanism is used, and the node listens to the channel before transmitting and transmits immediately when idle. When the node utilization rate is greater than the first node utilization rate threshold, the TDMA mechanism is used to allocate a fixed transmission time slot to each node. When the second node utilization threshold is reached, the dynamic scheduling unit automatically triggers protocol switching based on the real-time node utilization; the second node utilization threshold is greater than the first node utilization threshold.

2. The underground space mapping and environmental monitoring system according to claim 1, characterized in that, Sensor nodes are deployed in underground spaces using unmanned mobile devices to collect various parameter data, including environmental parameter data, spatial image data, and node status parameters.

3. The underground space mapping and environmental monitoring system according to claim 1, characterized in that, The edge computing unit first predicts the energy, bandwidth, and computing resource requirements of a task based on historical task data, node energy status, and real-time network status. The formula for calculating the node energy state ,in, This represents the remaining energy at time t; Represents the initial energy of the node; This indicates the energy consumption rate of the current node; This represents the time step, or the frequency of node energy updates. Finally, based on node load, traffic, and topology changes, the network congestion risk is predicted, and the congestion area and degree of congestion are output.

4. The underground space mapping and environmental monitoring system according to claim 1, characterized in that, The dynamic scheduling unit also includes a prediction and matching module, which specifically includes pre-deploying cluster head selection, path planning and task priority strategies based on the prediction results of the edge computing unit; Compare the deviation between the preset strategy and the actual execution. If the time delay deviation is less than or equal to the first time delay threshold and the energy consumption deviation is less than or equal to the first energy threshold, execute the preset strategy. When the time delay deviation is greater than the first time delay threshold and / or the energy consumption deviation is greater than the first energy threshold, a new cluster head is selected and the transmission path is updated by rerunning the PSO algorithm based on the current node status, network load and task priority.

5. The underground space mapping and environmental monitoring system according to claim 1, characterized in that, The simulation optimization unit verifies the scheduling strategy using multiple metrics, including simulating the network load of multiple nodes and testing end-to-end transmission latency; monitoring node energy consumption rates; and simulating high-traffic scenarios to test packet loss rate and throughput.

6. The underground space mapping and environmental monitoring system according to claim 1, characterized in that, It also includes an energy management unit, which is connected to the data acquisition unit and the dynamic scheduling unit; it includes adjusting the sensor node acquisition frequency; controlling the node transmission power; triggering the sleep mode of low-energy nodes; activating the energy supply of high-priority tasks when the task priority is reduced, and delaying the execution of low-priority tasks.

7. A method for underground space mapping and environmental monitoring, applicable to an underground space mapping and environmental monitoring system as described in any one of claims 1-6, characterized in that, Specifically, it includes: Step 1: Multiple sensor nodes acquire various parameter data; Step 2: The edge computing unit receives various parameter data and predicts future tasks, node energy consumption, and network congestion risks based on historical data and real-time status; it then uses machine learning algorithms and historical data analysis methods to generate a predictive strategy for task execution. Step 3: The dynamic scheduling unit selects sensor nodes based on future tasks, node energy consumption, and network congestion risks, aggregates the collected data from the sensor nodes, and executes a low-latency routing algorithm to dynamically adjust and obtain a scheduling strategy. The low-latency routing algorithm includes a clustering routing submodule, an opportunistic routing submodule, and a multi-objective optimization submodule. The clustering routing submodule divides the sensor nodes into multiple clusters and selects the globally optimal cluster head node based on the node's remaining energy, distance to the base station, and load using a particle swarm optimization algorithm. The opportunistic routing submodule monitors the node's remaining energy, load, and distance to the target node in real time and dynamically selects nodes with sufficient energy and low load as the next-hop transmission path. The multi-objective optimization submodule combines the minimum spanning tree algorithm with communication distance and energy consumption as dual objectives to optimize the transmission path between cluster heads and minimize the total transmission cost. Step 4: The simulation optimization unit verifies the performance of the scheduling strategy, and the verification results are fed back to the dynamic scheduling unit to optimize the particle velocity parameters of the PSO algorithm and the weight coefficients of MST. The data acquisition unit, edge computing unit, and dynamic scheduling unit communicate with each other through a wireless sensor network, using a hybrid CSMA / TDMA protocol. When the node utilization rate is less than or equal to the first node utilization rate threshold, the CSMA mechanism is used, and the node listens to the channel before transmitting and transmits immediately when idle. When the node utilization rate is greater than the first node utilization rate threshold, the TDMA mechanism is used to allocate a fixed transmission time slot to each node. When the second node utilization threshold is reached, the dynamic scheduling unit automatically triggers protocol switching based on the real-time node utilization; the second node utilization threshold is greater than the first node utilization threshold.

Images