Unmanned aerial vehicle data collection method and system based on region division and path optimization

Abstract

The invention provides an unmanned aerial vehicle data collection method and system based on region division and path optimization, and relates to the technical field of wireless communication networks, and the method comprises the steps: determining a communication reachable relation matrix according to the spatial position of a node in a wireless sensor network, and determining a network topology structure diagram according to the communication reachable relation matrix; according to the network topology structure diagram, the area of the wireless sensor network is divided into a plurality of sub-areas with the fact that the spatial distance between any two nodes in the same sub-area is smaller than a target threshold value and the connection relation exists between adjacent sub-areas as targets; determining an entrance point and an exit point of each sub-region according to a connection relationship between the adjacent sub-regions; planning a flight path from an entrance point to an exit point of a single sub-region and passing through nodes in the sub-region for the unmanned aerial vehicle; and planning a global flight path according to the flight paths of the plurality of sub-regions and the connection relationship between the adjacent sub-regions, and controlling the unmanned aerial vehicle to execute a data collection task along the global flight path. The invention aims to reduce the complexity of unmanned aerial vehicle data acquisition planning.

Description

A Method and System for UAV Data Collection Based on Region Division and Path Optimization Technical Field

[0001] This application relates to the field of wireless communication network technology, specifically to a method and system for collecting UAV data based on region division and path optimization. Background Technology

[0002] With the rapid development of the Internet of Things (IoT), its technology is widely applied in industries such as manufacturing, agriculture, and transportation. However, due to the random, dense, and low-power characteristics of IoT node deployment, efficient data collection schemes using single drones can improve network lifetime, connectivity, and data timeliness, thus showing promising application prospects. Current drone data collection technologies mainly fall into two categories: globally optimized and reinforcement learning-based. Globally optimized drone data collection methods are suitable for small to medium scales. When the number of nodes is large, the path planning problem becomes too complex, computationally expensive, and difficult to solve. Reinforcement learning-based drone data collection, on the other hand, often relies on the network topology used during training for its flight strategy. When migrating to new structures, training performance significantly degrades. Furthermore, for large-scale wireless sensor networks, the increase in the state space dimension can make it difficult for reinforcement learning algorithms to converge. Summary of the Invention

[0003] In view of this, this application provides a UAV data collection method based on region division and path optimization. It aims to solve or partially solve the problems existing in the background art.

[0004] The first aspect of this application provides a method for collecting UAV data based on region partitioning and path optimization. The method includes: determining a communication reachability matrix of the wireless sensor network based on the spatial location of each node in the wireless sensor network, wherein the elements in the communication reachability matrix... The value of represents whether the i-th node and the j-th node in the wireless sensor network can communicate directly. Based on the communication reachability matrix of the wireless sensor network, a network topology diagram of the wireless sensor network is determined. The nodes in the network topology diagram are the nodes in the wireless sensor network, and the edges in the network topology diagram are determined based on the communication reachability matrix. Based on the network topology diagram of the wireless sensor network, with the objective that the spatial distance between any two nodes in the same sub-region is less than a target threshold and there is a connection between adjacent sub-regions, the region of the wireless sensor network is divided into multiple sub-regions. Based on the connection between adjacent sub-regions in the multiple sub-regions, the entry point and exit point of each sub-region are determined. For each sub-region in the multiple sub-regions, a flight path is planned for the UAV from the entry point to the exit point of that sub-region, passing through the nodes in that sub-region. Based on the flight paths of each sub-region and the connection between adjacent sub-regions in the multiple sub-regions, a global flight path is planned for the UAV within the region of the wireless sensor network, and the UAV is controlled to perform data collection tasks according to the global flight path.

[0005] A second aspect of this application provides a UAV data collection system based on region partitioning and path optimization. The system includes a matrix determination module, used to determine a communication reachability matrix of the wireless sensor network based on the spatial location of each node in the wireless sensor network. The elements in the communication reachability matrix... The value of represents whether the i-th node and the j-th node in the wireless sensor network can communicate directly; the network topology determination module is used to determine the network topology diagram of the wireless sensor network based on the communication reachability matrix of the wireless sensor network, wherein the nodes in the network topology diagram are the nodes in the wireless sensor network, and the edges in the network topology diagram are determined based on the communication reachability matrix; the sub-region division module is used to divide the wireless sensor network into regions based on the network topology diagram of the wireless sensor network, with the objective that the spatial distance between any two nodes in the same sub-region is less than a target threshold and there is a connection relationship between adjacent sub-regions. The system is divided into multiple sub-regions; an entrance / exit determination module is used to determine the entrance and exit points of each sub-region based on the connection relationships between adjacent sub-regions; a flight path planning module is used to plan a flight path for the UAV from the entrance point to the exit point of each sub-region, passing through the nodes within that sub-region; and a global flight path planning module is used to plan a global flight path for the UAV within the wireless sensor network based on the flight paths of each sub-region and the connection relationships between adjacent sub-regions, and to control the UAV to perform data collection tasks according to the global flight path.

[0006] This application provides a UAV data collection method based on region partitioning and path optimization, which has the following advantages: This method guides UAVs in path planning within large-scale and randomized wireless sensor networks. The path planning process is divided into two layers: the upper layer partitions regions, ensuring each region has at least one connectable entry and exit point to accommodate continuous UAV access; the lower layer performs trajectory planning within each partitioned sub-region, enabling the UAV to achieve both high coverage and short path lengths. Attached Figure Description

[0007] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the description of the embodiments of this application will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0008] Figure 1 is a flowchart illustrating a UAV data collection method based on region division and path optimization according to an embodiment of this application; Figure 2 is a schematic diagram of sub-region division and path planning in a UAV data collection method based on region division and path optimization according to an embodiment of this application; Figure 3 is a schematic diagram of a UAV data collection system based on region division and path optimization according to an embodiment of this application. Detailed Implementation

[0009] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0010] Wireless sensor networks (WSNs) typically consist of sensor nodes, sink nodes, and an external network. Sensor nodes monitor their surrounding environment and generate data packets based on timed triggers or events. Sink nodes serve as the interface between the WSN and the external network and can be powerful infrastructure such as storage nodes or base stations. Data packets generated by sensor nodes are transmitted to the sink node via multi-hop routing. The distribution of nodes in a WSN is usually random, and the topology may dynamically change as node energy consumption increases. Drones are considered mobile sink nodes or external network exits in WSNs. Drones approach existing sensor nodes or sink nodes in the WSN by flying, establishing transmission links with nodes while passing by or hovering to collect data.

[0011] Referring to Figure 1, Figure 1 is a flowchart illustrating a UAV data collection method based on region partitioning and path optimization according to an embodiment of this application. This application specifically relates to a data collection method using a single UAV based on region partitioning and path optimization in a large-scale, irregularly structured wireless sensor network with random topology. As shown in Figure 1, the method includes: Step S01: Determining the communication reachability matrix of the wireless sensor network based on the spatial location of each node in the wireless sensor network, wherein the elements in the communication reachability matrix... The value of represents whether the i-th node and the j-th node in the wireless sensor network can communicate directly.

[0012] In this embodiment, the spatial range of the wireless sensor network for which UAV data collection is required is first determined, and the spatial location of each node in the wireless sensor network within this spatial range is obtained. The nodes in the wireless sensor network include at least one type of sensor node and at least one type of aggregation node. Sensor nodes are used to sense environmental data and generate data packets to be collected. These data packets are transmitted to the aggregation node via multi-hop routing, where the aggregation node temporarily stores the data for UAV data collection. The aggregation node receives data from one or more sensor nodes and serves as a data aggregation point. The spatial locations of the sensor nodes in the sensor node set can be obtained through their GPS systems, and the spatial locations of the aggregation nodes in the aggregation node set can be obtained through base station measurement technology. The spatial locations of the nodes are represented using two-dimensional or three-dimensional spatial coordinates, such as a node set. Where N represents the total number of nodes in the wireless sensor network. Indicates the first The spatial location of each node.

[0013] In this embodiment, network topology information of the wireless sensor network is obtained based on the spatial location of each node in the wireless sensor network. This information is used to characterize whether there is a direct data communication relationship between the nodes in the wireless sensor network. The network topology information is represented as a communication reachability matrix. One optional implementation for determining the communication reachability matrix is ​​as follows: a distance threshold is preset. When the spatial distance (e.g., Euclidean distance) between two nodes is less than or equal to this threshold, it is determined that the two nodes can communicate directly. In this case, the value of the matrix element corresponding to the two nodes in the communication reachability matrix is ​​set to 1. When the spatial distance (e.g., Euclidean distance) between two nodes is greater than this threshold, it is determined that the two nodes cannot communicate directly. In this case, the value of the matrix element corresponding to the two nodes in the communication reachability matrix is ​​set to 0. The specific expression is: ,in, This represents the Euclidean distance between the i-th node and the j-th node. This indicates that the i-th node and the j-th node can communicate directly. This indicates that the i-th node and the j-th node cannot communicate directly. Through this implementation, a communication reachability matrix corresponding to the wireless sensor network can be constructed, and the elements in this communication reachability matrix... The value of represents whether the i-th node and the j-th node in the wireless sensor network can communicate directly. A value of 1 indicates that direct communication is possible, and a value of 0 indicates that direct communication is not possible. The nodes in the communication reachability matrix include sensor nodes and aggregation nodes.

[0014] Step S02: Determine the network topology diagram of the wireless sensor network based on the communication reachability matrix of the wireless sensor network. The nodes in the network topology diagram are the nodes in the wireless sensor network, and the edges in the network topology diagram are determined based on the communication reachability matrix.

[0015] In this embodiment, the network topology information can be determined based on the communication reachability matrix, and represented as a network topology graph. Specifically, the nodes in the wireless sensor network are identified as nodes in the network topology graph, and the network topology graph is constructed. Simultaneously, based on the obtained communication reachability matrix of the wireless sensor network, an edge is connected to a pair of nodes in the network topology graph that can communicate directly with each other, thereby constructing the network topology graph of the wireless sensor network. The set of nodes in this network topology graph is the set of nodes in the wireless sensor network, and the set of edges in the network topology graph is determined based on the communication reachability matrix of the wireless sensor network. Ultimately, the wireless sensor network will be abstracted as a network topology graph. This graph structure is used to describe the data transmission relationships between nodes in the wireless sensor network, without limiting the actual flight path of the UAV to follow the edges in this graph structure.

[0016] In one alternative implementation, the node distribution information and network topology information can be obtained through any one or more of the following methods: pre-configured and stored by the wireless sensor network during the deployment phase; obtained in real time by UAVs through broadcast detection and node response; or uniformly distributed by the ground control system or edge computing nodes.

[0017] Step S03: Based on the network topology diagram of the wireless sensor network, with the goal of ensuring that the spatial distance between any two nodes in the same sub-region is less than a target threshold and that there is a connection between adjacent sub-regions, the wireless sensor network is divided into multiple sub-regions.

[0018] In this embodiment, after obtaining the network topology diagram of the wireless sensor network through step S02, if there is an edge between two nodes in the network topology diagram, it indicates that the two nodes can communicate directly. Based on the obtained network topology diagram of the wireless sensor network, the wireless sensor network is divided into multiple sub-regions with the goal that the spatial distance between any two nodes in the same sub-region is less than a target threshold and that there is a connection between adjacent sub-regions. That is, for any sub-region, the spatial distance between any two nodes in the sub-region must be less than the target threshold, and for adjacent sub-regions, there must be a connection between them. The existence of a connection between adjacent sub-regions means that the interval between adjacent sub-regions is relatively short, such as two adjacent sub-regions sharing nodes, or two adjacent sub-regions having overlapping spatial ranges, or two adjacent sub-regions sharing a boundary, etc. The target threshold is greater than the distance threshold used to determine whether nodes can communicate directly.

[0019] Step S04: Determine the entry point and exit point of each sub-region based on the connection relationship between adjacent sub-regions in the plurality of sub-regions.

[0020] In this embodiment, based on the multiple sub-regions of the wireless sensor network obtained by division, and based on the connection relationships between adjacent sub-regions, the entry point and exit point of each sub-region are determined. Through the same implementation method, the corresponding entry point and exit point are determined for each of the divided sub-regions.

[0021] Step S05: For each of the multiple sub-regions, plan a flight path for the UAV from the entrance point to the exit point of the sub-region and passing through the nodes in the sub-region.

[0022] In this embodiment, for any one of the obtained sub-regions, a flight path is planned for the UAV from the entry point to the exit point of that sub-region, passing through each node within that sub-region. Thus, a corresponding flight path can be planned for each sub-region within the area of ​​the wireless sensor network.

[0023] Step S06: Based on the flight paths of each of the multiple sub-regions and the connection relationships between adjacent sub-regions, plan a global flight path for the UAV within the area of ​​the wireless sensor network, and control the UAV to perform data collection tasks according to the global flight path.

[0024] In this embodiment, after obtaining the flight path of each sub-region, a global flight path is planned for the UAV within the wireless sensor network area based on the UAV's flight paths in these multiple sub-regions and the connectivity between adjacent sub-regions. This global flight path is generated by connecting the exit point of the corresponding sub-region with the entry point of the adjacent sub-region. Finally, the UAV is controlled to perform a data collection task according to this global flight path, which is obtained by connecting the flight paths of the multiple sub-regions in a planned order. Specifically, the global flight path consists of the following parts in sequence: the flight path from the UAV's takeoff point to the entry point of the first sub-region; the local flight path within each sub-region; the cross-regional flight path between adjacent sub-regions from the exit point to the entry point; and the flight path from the exit point of the last sub-region to the UAV's return point. Based on the order of accessing sub-regions, the flight paths within each sub-region are spliced ​​together to generate the global flight path of the UAV. Specifically, the entry point of the first sub-region is used as the global starting point of the UAV, and the flight path of that sub-region is used as the first segment of the global flight path. For two adjacent sub-regions, the exit point of the previous sub-region is connected to the entry point of the next sub-region, and the flight paths within each sub-region are spliced ​​together in sequence to form a continuous global flight path.

[0025] The UAV data collection method based on region segmentation and path optimization provided in this application can guide UAVs in path planning within large-scale and random wireless sensor networks. This application breaks down the path planning process into two layers: the upper layer performs region segmentation, ensuring each region has at least one connectable entry and exit point to accommodate continuous UAV access; the lower layer performs trajectory planning within each segmented sub-region, enabling the UAV to achieve both high coverage and short path lengths.

[0026] In conjunction with the above embodiments, in one implementation, this application also provides a UAV data collection method based on region partitioning and path optimization. In this UAV data collection method based on region partitioning and path optimization, step S03 may include: Step S03_1: Determining the degree matrix of the wireless sensor network according to the network topology diagram of the wireless sensor network, wherein an element in the degree matrix represents the number of other nodes in the wireless sensor network that can directly communicate with that node.

[0027] In this embodiment, based on any node in the obtained network topology graph of the wireless sensor network, the number of other nodes in the network topology that are connected to that node by an edge is determined. Using the same implementation method, the number of edge connections for each node in the network topology can be obtained. Based on the obtained number of edge connections for each node in the network topology, a degree matrix of the wireless sensor network is constructed. Each element in the degree matrix represents the number of other nodes in the wireless sensor network that can directly communicate with the node corresponding to that element. For example, if the number of other nodes that can directly communicate with the i-th node in the wireless sensor network is 5, then the value of the element in the i-th row and i-th column of the degree matrix of the wireless sensor network is 5.

[0028] Step S03_2: Based on the degree matrix and the communication reachability matrix, with the objective that the spatial distance between any two nodes in the same sub-region is less than a target threshold and there is a connection between adjacent sub-regions, the network topology graph is divided into multiple subgraphs, each subgraph corresponding to a sub-region; wherein, a connection between adjacent sub-regions is determined if at least one of the following conditions is met: there are shared nodes between the two sub-regions; the spatial ranges of the two sub-regions overlap, or there is a shared boundary between the two sub-regions; there is at least one pair of nodes in the two sub-regions, and the spatial distance between the pair of nodes is less than a preset cross-region flight threshold.

[0029] In this embodiment, firstly, it can be based on the degree matrix. and communication reachability matrix Determine the Laplace matrix , Then, calculate the Laplace matrix. The eigenvectors corresponding to the first m smallest eigenvalues ​​are determined. Eigenvectors corresponding to smaller eigenvalues ​​exhibit stronger smoothness under graph structure constraints; that is, their values ​​change less between nodes with larger edge weights, thus characterizing the low-frequency structure of the graph. Eigenvectors corresponding to larger eigenvalues, on the other hand, show higher-frequency oscillations in the graph, reflecting local or noise-dominated change patterns, and are therefore usually discarded in spectral clustering. Next, the eigenvectors corresponding to the first m smallest eigenvalues ​​are used for sub-region partitioning. Specifically, spectral spatial clustering (e.g., using K-means) can be performed on the eigenvectors corresponding to the first m smallest eigenvalues ​​to obtain multiple clusters. Each cluster corresponds to a sub-region, and the nodes within each cluster are the nodes within that sub-region.

[0030] For cases where one sub-region is adjacent to another, determine whether these two adjacent sub-regions are connected. Specifically, a connection is determined between adjacent sub-regions if at least one of the following conditions is met: the two sub-regions share a node; the spatial extent of the two sub-regions overlaps; or the two sub-regions share a boundary; or there is at least one pair of nodes in the two sub-regions, and the spatial distance between these pairs of nodes is less than a preset cross-region flight threshold. The connection between adjacent sub-regions is used to describe the portability of the UAV between different sub-regions.

[0031] If a subregion is connected to all its neighboring subregions, it is designated as the final subregion. If a subregion is not connected to one or more neighboring subregions, it can be re-partitioned based on the node centrality score, targeting nodes within the two or more adjacent, unconnected subregions. Specifically, for each node within the two or more adjacent, unconnected subregions, the centrality score is calculated, and then one or more nodes with the highest centrality scores are selected as the center node to construct the subregion. For any given subregion, nodes whose distance to the center node of that subregion is less than half a target threshold are assigned to that subregion. The centrality score of a node is calculated as follows: ,in, This represents the centrality score of the i-th node. This represents the degree value of the i-th node obtained from the degree matrix, where α is the weight. This represents the communication reachability relationship between the i-th node and the j-th node, obtained based on the communication reachability matrix. This represents the degree value of the j-th node obtained based on the degree matrix. In this calculation formula, only nodes that can directly communicate with the i-th node will affect the centrality score of the i-th node.

[0032] In this embodiment, for a node not covered by any sub-region, one possible implementation is to find a central node that can communicate directly with the node, or that can communicate via multi-hop routing, and then assign the node to the sub-region corresponding to the central node closest to the node among all the found central nodes. Another possible implementation is to assign the node to other sub-regions based on its centrality score. Specifically, if the distance between the node and the central node of another sub-region is less than half a target threshold, then the node is assigned to that other sub-region.

[0033] Step S03_3: For each sub-region, check whether each node in the sub-region has a node that can communicate directly with it. If the node does not have a node that can communicate directly with it in the sub-region, then based on the number of nodes that can communicate directly with it in other sub-regions, assign the node to the other sub-region with the most nodes that can communicate directly with it.

[0034] In this embodiment, for any sub-region obtained by division, it is checked whether each node in the sub-region has a node that can communicate directly with it. If a node in the sub-region cannot communicate directly with any node in the sub-region, then the region with the most nodes that can communicate directly with that node is determined, and the node is then assigned to the sub-region with the most nodes that can communicate directly with it. That is, the sub-region with the most nodes that can communicate directly with that node is determined as the target sub-region, and the node is then assigned to that target sub-region.

[0035] Step S03_4: For a node that cannot communicate directly with any other node, assign the node to the sub-region that is closest to its spatial location.

[0036] In this embodiment, if a node cannot communicate directly with any of its child nodes, it is considered an isolated node. In this case, the node is assigned to the sub-region closest to its spatial location. The region assignment does not presuppose that the wireless sensor network is fully connected; even when the wireless sensor network is fully connected, multiple sub-regions can still be generated. A graph traversal algorithm (such as BFS or DFS) is used to verify the connectivity of the node sets within each sub-region.

[0037] In conjunction with the above embodiments, in one implementation, this application also provides a UAV data collection method based on region partitioning and path optimization. In this UAV data collection method based on region partitioning and path optimization, step S04 may include: Step S04_1: Based on whether there is a connection between the i-th sub-region and the j-th sub-region among the plurality of sub-regions, a sub-region adjacency graph is generated. The i-th node and the j-th node in the sub-region adjacency graph correspond to the i-th sub-region and the j-th sub-region, respectively. If there is a connection between the i-th sub-region and the j-th sub-region, an edge is added between the i-th node and the j-th node in the sub-region adjacency graph.

[0038] In this embodiment, after dividing the region into multiple sub-regions, a sub-region adjacency graph is generated based on whether there is a connection between the i-th sub-region and the j-th sub-region. The i-th node and the j-th node in this sub-region adjacency graph correspond to the i-th sub-region and the j-th sub-region, respectively. If there is a connection between the i-th sub-region and the j-th sub-region, an edge is added between the i-th node and the j-th node in the sub-region adjacency graph.

[0039] Step S04_2: Based on the adjacency graph of the sub-regions, the take-off and return positions of the UAV, the spatial distribution of the multiple sub-regions, and the amount of data to be collected in each sub-region, determine the access order of the multiple sub-regions.

[0040] In this embodiment, after obtaining the sub-region adjacency graph through step S04_1, the sub-region adjacency graph records whether any two sub-regions are connected. In determining the order in which the UAV visits all sub-regions, firstly, based on the UAV's takeoff and return positions, it is determined which sub-region is closest to the takeoff position. The sub-region closest to the takeoff position is the first sub-region in the access order, and the sub-region closest to the return position is the last sub-region in the access order. After determining the first sub-region in the access order, the sub-regions connected to the first sub-region are determined based on the sub-region adjacency graph. The sub-regions connected to the first sub-region will be considered as the next candidate sub-regions in the access order. At this point, the spatial distribution and the amount of data to be collected for each sub-region connected to the first sub-region are further determined. Preferably, the sub-region with a larger amount of data to be collected and a smaller spatial distribution is the next sub-region to be accessed. After obtaining the second sub-region to be visited, the next sub-region to be visited is determined based on the connectivity between sub-regions, the spatial distribution of connected sub-regions, and the amount of data to be collected. This process continues until all sub-regions are added to the visit sequence, resulting in the final visit order. This visit order can be a pre-calculated fixed order, or it can be dynamically adjusted during mission execution based on the remaining time or energy of the drone. For example, if the drone has limited remaining time or energy, making it impossible to visit all sub-regions, some sub-regions can be deleted.

[0041] Step S04_3: Determine the preceding and subsequent sub-regions of each sub-region according to the access order of the multiple sub-regions.

[0042] In this embodiment, after obtaining the access order of the UAV to each sub-region through step S04_2, the preceding and following sub-regions of each sub-region in the access order are determined. The preceding sub-region of a sub-region refers to the sub-region most recently needed to be accessed before visiting the current sub-region, according to the access order of the UAV to each sub-region obtained through step S04_2. The following sub-region of a sub-region refers to the sub-region most recently needed to be accessed after visiting the current sub-region, according to the access order of the UAV to each sub-region obtained through step S04_2.

[0043] Step S04_4: Based on the nodes within each sub-region and the preceding and subsequent sub-regions of that sub-region, determine the entry and exit points of that sub-region as nodes that satisfy at least one of the following conditions: located within the overlapping space or shared boundary between the sub-region and its preceding sub-region, and located within the overlapping space or shared boundary between the sub-region and its subsequent sub-region; having the smallest distance to the entry point of the preceding sub-region, and having the smallest distance to the entry point of the subsequent sub-region; being a convergence node within the sub-region, where a convergence node is used to aggregate data generated by multiple sensor nodes; and being a cluster center obtained by clustering the sensor nodes within the sub-region according to their spatial location.

[0044] In this embodiment, based on the nodes within each sub-region and the preceding and subsequent sub-regions of that sub-region, the nodes within that sub-region that satisfy at least one of the following conditions are determined as the entry and exit points of that sub-region: nodes located within the overlapping space or on the shared boundary between the sub-region and its preceding sub-region, and nodes located within the overlapping space or on the shared boundary between the sub-region and its subsequent sub-region; the node with the smallest distance to the entry point of the preceding sub-region, and the node with the smallest distance to the entry point of the subsequent sub-region; a convergence node within the sub-region, used to aggregate data generated by multiple sensor nodes; and a cluster center obtained by clustering the sensor nodes within the sub-region according to their spatial locations.

[0045] Nodes located within the overlapping space or shared boundary between a sub-region and its preceding sub-region can be designated as entry points or exit points of that sub-region. Similarly, the node with the smallest distance to the entry point of its preceding sub-region can be designated as the entry point, and the node with the smallest distance to the entry point of its subsequent sub-region can be designated as the exit point. A convergence node within the sub-region can also be designated as the entry point or exit point. Furthermore, the cluster center obtained by clustering the sensor nodes within the sub-region according to their spatial locations can be designated as the entry point or exit point. In this embodiment, for sub-regions located at opposite ends of the access sequence, the entry or exit point can coincide with the UAV's takeoff or return point.

[0046] In conjunction with the above embodiments, in one implementation, this application also provides a UAV data collection method based on region partitioning and path optimization. In this UAV data collection method based on region partitioning and path optimization, step S05 may include: Step S05_1: For the k-th sub-region, construct a k-th candidate access node set, wherein the nodes in the k-th candidate access node set include: each aggregation node within the k-th sub-region, cluster centers obtained by clustering each sensor node within the k-th sub-region according to its spatial location, and sensor nodes within the k-th sub-region whose real-time state meets a preset state threshold.

[0047] In this embodiment, for the k-th sub-region, a candidate access node set (i.e., the k-th candidate access node set) is constructed. The nodes in this candidate access node set include: each aggregation node within the k-th sub-region, sensor nodes within the k-th sub-region whose relevant real-time status meets a preset state threshold, and each cluster center obtained by geographic density clustering of each sensor node within the k-th sub-region according to its spatial location. Through the same implementation method, each sub-region can construct its own corresponding candidate access node set. Through the same implementation method, a corresponding candidate access node set can be constructed for each sub-region. Among them, sensor nodes whose relevant real-time status meets the preset state threshold include sensor nodes with remaining energy lower than a preset energy threshold and sensor nodes with data volume exceeding a preset data volume threshold.

[0048] Step S05_2: For the i-th candidate access node in the k-th candidate access node set, determine the access benefit value of the candidate access node based on the set of sensor nodes within the data collection range of the UAV around the candidate access node.

[0049] In this embodiment, for the i-th candidate access node in the k-th candidate access node set, the access benefit value of the i-th candidate access node is determined based on the set of sensor nodes within the drone data collection range surrounding the i-th candidate access node. The drone data collection range is the maximum spatial range within which the drone can receive data after hovering at a point, such as a radius of 50 meters. An optional implementation for determining the access benefit value of the i-th candidate access node is as follows: determine the set of sensor nodes within the drone data collection range surrounding the i-th candidate access node's access point; determine the amount of data to be collected for each sensor node in the set; and sum the amounts of data to be collected for each sensor node in the set to obtain the access benefit value of the i-th candidate access node. Preferably, when summing the amounts of data to be collected for each sensor node in the set, a weighted summation method can be used to calculate the access benefit value of the i-th candidate access node. For higher-priority sensor nodes, a higher weight is set when weighting the amounts of data to be collected for those sensor nodes. The specific expression is as follows: ,in, This represents the set of sensor nodes within the data collection range of the i-th access node in the k-th candidate access node set. It is the amount of data to be collected from the j-th sensor node in the sensor node set. These are the corresponding weighting coefficients, used to reflect the importance or priority of sensor nodes. Using the same implementation method, a corresponding access benefit value can be calculated for each candidate access node in each divided set of candidate access nodes. This access benefit value characterizes the amount of data that the UAV can obtain by collecting data at that candidate access point.

[0050] Step S05_3: Based on the access revenue value of each candidate access node in the k-th candidate access node set, select a portion of candidate access nodes from the k-th candidate access node set to obtain the k-th access node set.

[0051] In this embodiment, after calculating the access benefit value of each candidate access node in the k-th candidate access node set, the candidate access nodes in the k-th candidate access node set are sorted from highest to lowest access benefit value. Then, a certain number of candidate access nodes with high ranking are selected from the k-th candidate access node set, and these selected candidate access nodes form the k-th access node set corresponding to the k-th sub-region. The selected candidate access nodes, as part of the access node set, are referred to as access nodes. Using the same implementation method, a corresponding access node set can be calculated for each sub-region.

[0052] Step S05_4: For the i-th access node in the k-th access node set, determine the hovering time of the access node based on the set of sensor nodes within the data collection range of the UAV around the access node.

[0053] In this embodiment, after obtaining the set of the kth access nodes corresponding to the kth sub-region through step S05_3, for the i-th access node in the set of the kth access nodes, the hovering time of the i-th access node is determined based on the set of sensor nodes within the data collection range of the UAV surrounding the i-th access node and the total amount of data to be collected corresponding to the set of sensor nodes. The hovering time represents the minimum dwell time required for the UAV to complete the data collection of the corresponding access node. The specific calculation expression is as follows: ,in, Let be the hover time of the i-th access node corresponding to the k-th sub-region. This indicates the effective data transmission rate between the drone and the node; This represents the total amount of data to be collected from the set of sensor nodes within the data collection range of the i-th access node in the k-th access node set.

[0054] Step S05_5: Based on the hovering time of each access node in the k-th access node set, plan a flight path for the UAV from the entrance point to the exit point of the sub-region, passing through the nodes in the sub-region.

[0055] In this embodiment, the flight cost is determined to be hovering time. Based on the hovering time of each access node in the set of the kth access nodes corresponding to the kth sub-region, a flight path with minimum flight cost is planned for the UAV from the entry point to the exit point of the sub-region, passing through nodes within the sub-region. Any heuristic path planning method can be used to plan a flight path with minimum flight cost for the sub-region, with the goal of traveling from the entry point to the exit point and passing through nodes within the sub-region.

[0056] In this embodiment, the specific method for the UAV to collect data along the global flight path is as follows: the UAV is controlled to sequentially enter each sub-region. Within any sub-region, it flies according to the generated flight path for that sub-region. When the UAV reaches an access node, it performs a hovering operation and receives the data cached by the corresponding aggregation node or sensor node. After completing the hovering time, it continues to perform subsequent flight tasks. After completing the data collection for the current sub-region, it flies out from the exit point of that sub-region and enters the next sub-region. After completing the data collection task, the data collection results of the UAV are output. As a specific embodiment, the results output by the UAV include the actual set of sub-regions visited and the access order, the set of access nodes in each sub-region, the cumulative flight time and cumulative hovering time of the UAV, and the total amount of data actually collected.

[0057] In conjunction with the above embodiments, in one implementation, this application also provides a UAV data collection method based on region partitioning and path optimization. This UAV data collection method based on region partitioning and path optimization further includes: using the i-th and j-th access nodes in the k-th access node set as path nodes in the candidate flight path, determining the flight time and energy consumption required for the UAV to fly between the i-th and j-th access nodes, thereby determining the flight cost of the UAV flying between the i-th and j-th access nodes.

[0058] In this embodiment, in another optional implementation, the flight cost includes both flight time and flight energy consumption. When the flight cost includes both flight time and flight energy consumption, the flight cost is calculated as follows: The entry point and exit point of the k-th sub-region are added to the k-th access node set for flight cost calculation. The i-th and j-th access nodes in the k-th access node set corresponding to the k-th sub-region are used as path nodes in the candidate flight path. The flight time and flight energy consumption required for the UAV to fly between the i-th and j-th access nodes are determined to determine the flight cost of the UAV flying between the i-th and j-th access nodes.

[0059] One possible implementation for determining the flight cost of a UAV flying between the i-th and j-th access nodes is to directly add the flight time and energy consumption of the UAV between the i-th and j-th access nodes to obtain the corresponding flight cost. Another possible implementation is to use a weighted summation of the flight time and energy consumption of the UAV between the i-th and j-th access nodes to obtain the corresponding flight cost, specifically expressed as follows: ,in, This indicates that the drone flies from the i-th access point in the set of k-th access nodes corresponding to the k-th sub-region to the j-th access point. Flight time, This represents the flight energy consumption during the corresponding flight process, with α and β being the corresponding weighting coefficients. The UAV's flight energy consumption is affected by flight altitude and speed (the higher the speed, the greater the wind resistance to overcome), while the UAV's flight time is affected by distance and speed. The flight time of the UAV between the i-th and j-th access nodes consists of the sum of the actual flight time and the hovering time. If the UAV flies from the i-th access node to the j-th access node, then the flight time from the i-th access node to the j-th access node includes the hovering time at the i-th access node plus the actual flight time from the i-th access node to the j-th access node. For example, if the UAV hovers at the i-th access node for 20 seconds and then takes 1 minute to fly from the i-th access node to the j-th access node, then the flight time from the i-th access node to the j-th access node is determined to be 1 minute and 20 seconds. The flight energy consumption of the UAV between the i-th and j-th access nodes consists of the sum of the actual flight energy consumption and the hovering energy consumption. If the UAV flies from the i-th access node to the j-th access node, the flight energy consumption of the UAV from the i-th access node to the j-th access node includes the hovering energy consumption of the UAV at the i-th access node plus the actual flight energy consumption of the UAV from the i-th access node to the j-th access node.

[0060] In this embodiment, the flight cost between any two access nodes in each set of access nodes can be calculated using the same implementation method.

[0061] In this application, if the method further includes determining the flight cost based on flight time and flight energy consumption, step S05_5 may include: step S05_5a: using the flight cost of the UAV flying between any two access nodes in the kth access node set as the edge weight, and using each access node in the kth access node set as a node, construct the kth directed graph.

[0062] In this embodiment, for any sub-region obtained by partitioning, an Orienteering graph model with start and end constraints is constructed within that sub-region. In this Orienteering graph model, the start point is the entry point of the sub-region, and the end point is the exit point of the sub-region. Using the same implementation method, each sub-region can construct its own corresponding Orienteering graph model with start and end constraints. The Orienteering graph model of a single sub-region allows the UAV to access only a portion of the access nodes in the set of access nodes corresponding to that sub-region. Since the implementation method for constructing the corresponding Orienteering graph model with start and end constraints for each sub-region is the same, we will use one sub-region as an example for explanation, representing it as the k-th sub-region. An Orienteering graph model with start and end constraints is constructed within the k-th sub-region, and this Orienteering graph model is represented as a directed graph. The edge weights represent the flight cost of the UAV flying between two access points (including both flight time and energy consumption, as mentioned above). The nodes in this directed graph are composed of access nodes from the set of access nodes corresponding to the kth sub-region. Specifically, the directed graph is constructed as follows: using the flight cost of the UAV flying between any two access nodes in the set of access nodes corresponding to the kth sub-region as the edge weights, and simultaneously using each access node in that set as a node, a kth directed graph corresponding to the kth sub-region is constructed. Through the same implementation method, a directed graph corresponding to each sub-region can be constructed.

[0063] Step S05_5b: Under the condition of satisfying the k-th constraint, with the goal of maximizing the total access revenue, determine whether each access node in the k-th access node set is accessed to obtain the flight path of the sub-region. The k-th constraint includes: the sum of the total flight time and the total hovering time of the UAV in the k-th sub-region does not exceed a preset time; the sum of the total flight energy consumption and the total hovering energy consumption of the UAV in the k-th sub-region does not exceed a preset energy; the flight path of the UAV starts from the entrance point of the k-th sub-region and ends at the exit point of the k-th sub-region.

[0064] In this embodiment, after obtaining the Orienteering graph model for each sub-region through step S05_5a, the Orienteering graph model for each sub-region is optimized using the same model optimization objective. The model optimization objective is to maximize the total access revenue for each sub-region. The specific model optimization expression is as follows: ,in, Let k represent the set of the visited nodes. This represents the access reward value of the i-th access node in the k-th access node set. This indicates whether the i-th access node in a planned candidate flight path for the k-th sub-region has been visited.

[0065] In this embodiment, for any sub-region among all the divided sub-regions, when determining a flight path corresponding to that sub-region with the objective of maximizing the total access revenue of that sub-region, the following constraints must be met: the sum of the total flight time and total hovering time of the UAV in that sub-region does not exceed a preset time; the sum of the total flight energy consumption and total hovering energy consumption of the UAV in that sub-region does not exceed a preset energy; the UAV's flight path starts from the entry point of that sub-region and ends at the exit point of that sub-region. The edge weights in the directed graph corresponding to the sub-region are used to determine whether the total flight cost of a candidate flight path (which includes the total actual flight time, total hovering time, and total energy consumption of the candidate flight path) satisfies the constraints set for the flight path. That is, for each sub-region, under the constraints of its corresponding entry and exit points, a flight path for the UAV within that sub-region is generated to complete as much data collection as possible under limited flight time and energy constraints. As shown in Figure 2, the flight path planned by this application for a single sub-region is a path that maximizes the total access revenue under set constraints. The UAV will fly along this flight path in the sub-region and hover when it reaches each access node to collect data from the nodes around that access node.

[0066] In this embodiment, the flight path of the UAV within each sub-region is generated by solving the Orienteering graph model. In an optional implementation, before or during the execution of the UAV entering the corresponding sub-region, the candidate access point set, access value weight, and hovering time allocation are dynamically adjusted based on the changes in the data generation rate within that sub-region to improve the real-time performance and effectiveness of data collection. This application divides a large-scale random network into multiple sub-regions of controlled size by forcibly scaling down the wireless sensor network. Even when the network is a fully connected graph, it can still stably generate multiple sub-regions, structurally avoiding the problem of existing technologies being unable to directly handle large-scale random topologies, thus improving the feasibility and stability of the method.

[0067] Currently, for the problem of data collection by UAVs in wireless sensor networks, decision-making methods based on global path planning or reinforcement learning are commonly used. These methods usually rely on the assumption of relative stability of network topology, number of nodes, and spatial distribution. When the network size increases, node positions change randomly, or the network is not fully connected, problems such as model failure, unstable planning results, or a sharp increase in computational complexity can easily occur.

[0068] Compared to current methods that use reinforcement learning or deep learning to plan UAV flight paths, these methods typically require extensive training under specific network sizes, node distributions, or mission scenarios. Their performance is highly dependent on the distribution of training data, and generalization performance is prone to decline once the network topology, number of nodes, or data generation patterns change. This application, however, does not employ a learning-based decision model that relies on offline training or scenario fitting. Instead, it uses a path planning framework based on graph structure analysis and rule-driven approaches. Path generation is achieved through region partitioning, entry / exit constraints, and an orienteering graph model. This makes the method naturally adaptable to changes in the number of nodes, the randomness of spatial distribution, and network connectivity, significantly improving generalization capabilities across different wireless sensor network scenarios.

[0069] Compared to existing methods that directly perform global path planning for the entire wireless sensor network, the computational complexity increases exponentially or exponentially with the number of nodes, making it difficult to meet the real-time or near-real-time requirements of practical systems. This application, however, decomposes the global problem into multiple region-level sub-problems and performs path planning only within a single sub-region. This structurally transforms the original high-complexity global problem into multiple locally sized problems, effectively reducing the overall computational complexity and making the method more suitable for deployment on resource-constrained UAV platforms or edge computing environments.

[0070] In conjunction with the above embodiments, in one implementation, this application also provides a UAV data collection method based on region partitioning and path optimization. In this UAV data collection method based on region partitioning and path optimization, step S05_3 may include: selecting candidate access nodes whose access revenue value is greater than a preset threshold; or sorting each candidate access node in the k-th candidate access node set in descending order of access revenue value, and selecting a subset of candidate access nodes based on the sorting result; or determining the access cost-effectiveness of a candidate access node based on the access revenue value of each candidate access node in the k-th candidate access node set and the total flight cost of the UAV flying from that candidate access node to other candidate access nodes, sorting each candidate access node in the k-th candidate access node set in descending order of access cost-effectiveness, and selecting a subset of candidate access nodes based on the sorting result.

[0071] In this embodiment, to enable the UAV to collect more and more important data with limited resources based on the selected access nodes, this application provides another optional screening implementation method. Specifically, its core is to select a portion of candidate access nodes from the set of candidate access nodes corresponding to the kth sub-region to form the set of access nodes corresponding to the kth sub-region, under the premise of satisfying the UAV resource constraints. The resource constraints set in this application include at least one or more of the following: the maximum flight time constraint of the UAV in a single sub-region; the maximum available energy constraint of the UAV in a single sub-region; and the maximum allowed task execution time of the UAV in the sub-region. The screening rules for candidate access nodes set in this application include any one of the following three: 1. Select candidate access nodes with access benefit values ​​greater than a preset threshold as access nodes to form the access node set.

[0072] 2. Sort the candidate access nodes in the k-th candidate access node set in descending order of access revenue value, and select a certain number of the top-ranked candidate access nodes as access nodes to form the k-th access node set.

[0073] 3. Based on the access benefit value of each candidate access node in the k-th candidate access node set, and the sum of the flight costs for the UAV to fly from that candidate access node to all other candidate access nodes, a total flight cost is obtained. Based on the total flight cost and access benefit value of the candidate access node, an access cost-effectiveness ratio is determined for that candidate access node. This cost-effectiveness ratio can be selected as the result of dividing the access benefit value of the candidate access node by the total flight cost of the candidate access node. Finally, the candidate access nodes in the k-th candidate access node set are sorted in descending order of access cost-effectiveness ratio. Based on the sorting result, a certain number of candidate access nodes at the top of the sorting are selected as access nodes to form the k-th access node set.

[0074] Using the same implementation method, each sub-region can select a portion of candidate access nodes from its own candidate access node set to form its own access node set.

[0075] In conjunction with the above embodiments, in one implementation, this application also provides a UAV data collection method based on region partitioning and path optimization. In this UAV data collection method based on region partitioning and path optimization, before step S06, the method further includes: step S001: verifying whether the global flight path meets global constraints, the global constraints including at least one or more of the following: the total flight time of the UAV flying according to the global flight path does not exceed the maximum usable time of the UAV; the sum of the total flight energy consumption and the total hovering energy consumption of the UAV flying according to the global flight path does not exceed the maximum usable energy of the UAV; the sum of the total flight time and the total hovering time of the UAV flying according to the global flight path does not exceed the maximum execution time allowed for the data collection task.

[0076] In this embodiment, to ensure the security of the UAV data acquisition process, this application pre-sets global constraints for the global flight path. These global constraints are used to evaluate whether the UAV's resources are sufficient to complete the data acquisition task corresponding to the planned global flight path. Specifically, the planned global flight path is checked to ensure it meets the global constraints, which include at least one or more of the following: the total actual flight time required for the UAV to fly along the global flight path (excluding hovering time) does not exceed the UAV's maximum usable time; the sum of the total flight energy consumption and total hovering energy consumption of the UAV flying along the global flight path does not exceed the UAV's maximum usable energy; and the sum of the total flight time and total hovering time of the UAV flying along the global flight path does not exceed the maximum allowed execution time of the data collection task.

[0077] Step S002: If the global flight path is detected to not satisfy the global constraint, perform at least one of the following adjustments until a global flight path that satisfies the global constraint is obtained: delete at least one sub-region among the plurality of sub-regions; reduce the set of access nodes for at least one sub-region among the plurality of sub-regions; adjust the access order of the plurality of sub-regions. In this embodiment, if the planned global flight path does not satisfy the global constraint, adaptive adjustments are made to ensure that the re-determined global flight path after adjustment can satisfy the global constraint. Among them, at least one of the following adjustments is performed: delete at least one sub-region among all the sub-regions obtained by dividing the wireless sensor network. Data will not be collected from the deleted sub-regions. This method can effectively reduce the execution time and energy consumption of a single UAV when performing a data collection task; reduce the set of access nodes for at least one sub-region among all the sub-regions obtained by dividing the network. The reduction of access nodes can also effectively reduce the execution time and energy consumption of a single UAV when performing a data collection task; adjust the access order of all the sub-regions obtained by dividing the network.

[0078] In this application, the UAV data collection method based on region division and path optimization provided in this application further includes steps S001 and S002, controlling the UAV to perform the data collection task according to the global flight path, including: controlling the UAV to perform the data collection task according to the global flight path that satisfies the global constraints.

[0079] In this embodiment, if the planned global flight path satisfies the global constraint, the UAV is controlled to perform the data collection task according to the global flight path that satisfies the global constraint.

[0080] In conjunction with the above embodiments, in one implementation, this application also provides a UAV data collection method based on region partitioning and path optimization. In this UAV data collection method based on region partitioning and path optimization, the method further includes: during the UAV's data collection task, based on the UAV's remaining usable time or remaining usable energy, performing at least one of the following adjustments: deleting at least one unvisited sub-region from the plurality of sub-regions; reducing the set of access nodes for at least one unvisited sub-region from the plurality of sub-regions; and adjusting the access order of the unvisited sub-regions from the plurality of sub-regions.

[0081] In this embodiment, when planning the flight path for the UAV to perform data collection tasks, the hovering time of each access node, the actual flight time from one access node to another, and the energy consumption are all calculated and planned based on ideal theoretical conditions. However, in the actual data collection process, it is affected by many factors, such as prolonged headwind flight which consumes more energy, and the time required to travel from one access node to another in headwind conditions. Therefore, to ensure that the UAV can complete data collection safely, this application will perform at least one of the following adjustments during the UAV's data collection task, based on the UAV's remaining usable time or remaining usable energy: deleting at least one unvisited sub-region from all the divided sub-regions; reducing the set of access nodes for at least one unvisited sub-region from all the divided sub-regions; and adjusting the access order of the unvisited sub-regions from all the divided sub-regions. This adjustment can prevent the UAV from prematurely exhausting its resources and making an emergency landing if it is forced to complete the data collection task corresponding to the global flight path, which could also lead to the risk of the UAV crashing.

[0082] In conjunction with the above embodiments, in one implementation, this application also provides a UAV data collection method based on region division and path optimization. This UAV data collection method based on region division and path optimization further includes: when the UAV reaches the entry point of any of the plurality of sub-regions, re-acquiring the access revenue value of each access node in the access node set of that sub-region, and determining the change in the access revenue value of each access node in that sub-region; if the change in the access revenue value of at least one access node in that sub-region exceeds a preset change amount, without changing the entry point and exit point of that sub-region, replanning a flight path for the UAV from the entry point to the exit point of that sub-region, passing through the nodes in that sub-region, according to the re-acquired access revenue value.

[0083] In this embodiment, to enable the UAV to better adapt to the data acquisition task in real time, when the UAV reaches the entry point of any of the divided sub-regions, the application re-acquires the access revenue value of each access node in the access node set of that sub-region and determines the change in the access revenue value of each access node in that sub-region. If the change in the access revenue value of at least one access node in that sub-region exceeds a preset change amount, it is determined that there is a significant change in the amount of data in that sub-region compared to the initially uniformly determined situation. At this time, without changing the entry and exit points of the sub-region, the UAV is re-planned according to the re-acquired access revenue values, taking a new flight path from the entry point to the exit point of the sub-region and passing through the nodes in that sub-region. This allows for adaptive adjustment of the UAV's acquisition strategy based on the changes in the amount of data at each node in the sub-region, thereby adaptively acquiring more and better data for that sub-region.

[0084] Based on the same inventive concept, this application provides a UAV data collection system based on region division and path optimization, as shown in Figure 3. The system 300 includes: a matrix determination module 301, used to determine the communication reachability matrix of the wireless sensor network based on the spatial location of each node in the wireless sensor network, wherein the elements in the communication reachability matrix... The value of represents whether the i-th node and the j-th node in the wireless sensor network can communicate directly; the network topology determination module 302 is used to determine the network topology diagram of the wireless sensor network according to the communication reachability matrix of the wireless sensor network, wherein the nodes in the network topology diagram are the nodes in the wireless sensor network, and the edges in the network topology diagram are determined according to the communication reachability matrix; the sub-region division module 303 is used to divide the wireless sensor network into regions according to the network topology diagram of the wireless sensor network, with the objective that the spatial distance between any two nodes in the same sub-region is less than a target threshold and there is a connection relationship between adjacent sub-regions. The system comprises multiple sub-regions; an entrance / exit determination module 304, used to determine the entrance and exit points of each sub-region based on the connection relationships between adjacent sub-regions; a flight path planning module 305, used to plan a flight path for the UAV from the entrance point to the exit point of each sub-region, passing through nodes within that sub-region, for each sub-region; and a global flight path planning module 306, used to plan a global flight path for the UAV within the wireless sensor network based on the flight paths of each sub-region and the connection relationships between adjacent sub-regions, and to control the UAV to perform data collection tasks according to the global flight path.

[0085] It should be noted that, for the sake of simplicity, the method embodiments are all described as a series of actions. However, those skilled in the art should understand that the embodiments of this application are not limited to the described order of actions, because according to the embodiments of this application, some steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also understand that the embodiments described in the specification are all preferred embodiments, and the actions involved are not necessarily necessary for the embodiments of this application.

[0086] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0087] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, embodiments of this application can take the form of entirely hardware embodiments, entirely software embodiments, or embodiments combining software and hardware aspects. Furthermore, embodiments of this application can take the form of computer program products implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0088] This application describes embodiments with reference to flowchart illustrations and / or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing terminal device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal device, create means for implementing the functions specified in one or more blocks of the flowchart illustrations and / or one or more blocks of the block diagrams.

[0089] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing terminal device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means that implement the functions specified in one or more flowcharts and / or one or more block diagrams.

[0090] These computer program instructions may also be loaded onto a computer or other programmable data processing terminal equipment to cause a series of operational steps to be performed on the computer or other programmable terminal equipment to produce a computer-implemented process, such that the instructions, which execute on the computer or other programmable terminal equipment, provide steps for implementing the functions specified in one or more flowcharts and / or one or more block diagrams.

[0091] Although preferred embodiments of the present application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the embodiments of the present application.

[0092] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal device. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal device that includes said element.

[0093] The above provides a detailed description of the UAV data collection method and system based on region division and path optimization provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The description of the above embodiments is only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for collecting UAV data based on region division and path optimization, characterized in that, The method includes: determining a communication reachability matrix of the wireless sensor network based on the spatial location of each node in the wireless sensor network, wherein the elements in the communication reachability matrix... The value of represents whether the i-th node and the j-th node in the wireless sensor network can communicate directly. Based on the communication reachability matrix of the wireless sensor network, a network topology diagram of the wireless sensor network is determined. The nodes in the network topology diagram are the nodes in the wireless sensor network, and the edges in the network topology diagram are determined based on the communication reachability matrix. Based on the network topology diagram of the wireless sensor network, with the objective that the spatial distance between any two nodes in the same sub-region is less than a target threshold and there is a connection between adjacent sub-regions, the region of the wireless sensor network is divided into multiple sub-regions. Based on the connection between adjacent sub-regions in the multiple sub-regions, the entry point and exit point of each sub-region are determined. For each sub-region in the multiple sub-regions, a flight path is planned for the UAV from the entry point to the exit point of that sub-region, passing through the nodes in that sub-region. Based on the flight paths of each sub-region and the connection between adjacent sub-regions in the multiple sub-regions, a global flight path is planned for the UAV within the region of the wireless sensor network, and the UAV is controlled to perform data collection tasks according to the global flight path.

2. The UAV data collection method based on region division and path optimization according to claim 1, characterized in that, Based on the network topology diagram of the wireless sensor network, and with the objective that the spatial distance between any two nodes within the same sub-region is less than a target threshold and that there is a connection between adjacent sub-regions, the wireless sensor network is divided into multiple sub-regions. This includes: determining the degree matrix of the wireless sensor network based on the network topology diagram, where each element in the degree matrix represents the number of other nodes in the wireless sensor network that can directly communicate with that node; dividing the network topology diagram into multiple subgraphs based on the degree matrix and the communication reachability matrix, with the objective that the spatial distance between any two nodes within the same sub-region is less than a target threshold and that there is a connection between adjacent sub-regions, each subgraph corresponding to a sub-region; and for each sub-region, verifying the... If a node does not have a node that can communicate directly with it within a sub-region, then based on the number of nodes that can communicate directly with it in other sub-regions, the node is assigned to the sub-region with the most nodes that can communicate directly with it. For a node that cannot communicate directly with any node, the node is assigned to the sub-region that is spatially closest to it. A connection is determined between adjacent sub-regions if at least one of the following conditions is met: the two sub-regions share nodes; the spatial extent of the two sub-regions overlaps; or, the two sub-regions share a boundary; or there is at least one pair of nodes in the two sub-regions, and the spatial distance between the pair of nodes is less than a preset cross-region flight threshold.

3. The UAV data collection method based on region division and path optimization according to claim 1, characterized in that, Based on the connection relationships between adjacent sub-regions within the plurality of sub-regions, the entry and exit points of each sub-region are determined, including: generating a sub-region adjacency graph based on whether there is a connection relationship between the i-th sub-region and the j-th sub-region, where the i-th node and the j-th node in the sub-region adjacency graph correspond to the i-th sub-region and the j-th sub-region, respectively; if there is a connection relationship between the i-th sub-region and the j-th sub-region, an edge is added between the i-th node and the j-th node in the sub-region adjacency graph; determining the access order of the plurality of sub-regions based on the sub-region adjacency graph, the take-off and return positions of the UAV, the spatial distribution of the plurality of sub-regions, and the amount of data to be collected in each sub-region; and determining the entry and exit points of each sub-region based on the access order of the plurality of sub-regions. Each subregion is further divided into preceding and subsequent subregions. Based on each node within a subregion and its preceding and subsequent subregions, nodes within that subregion that satisfy at least one of the following conditions are identified as the entry and exit points of that subregion: located within the overlapping space or on the shared boundary between the subregion and its preceding subregion, and located within the overlapping space or on the shared boundary between the subregion and its subsequent subregion; having the smallest distance to the entry point of the preceding subregion, and having the smallest distance to the entry point of the subsequent subregion; being a convergence node within the subregion, where a convergence node is used to aggregate data generated by multiple sensor nodes; and being the cluster center obtained by clustering the sensor nodes within the subregion according to their spatial locations.

4. The UAV data collection method based on region division and path optimization according to claim 1, characterized in that, For each of the multiple sub-regions, a flight path is planned for the UAV from the entry point to the exit point of that sub-region, passing through the nodes within that sub-region. This includes: for the k-th sub-region, constructing a set of k-th candidate access nodes, wherein the nodes in the set of k-th candidate access nodes include: each convergence node in the k-th sub-region, cluster centers obtained by clustering each sensor node in the k-th sub-region according to its spatial location, and sensor nodes in the k-th sub-region whose real-time state meets a preset state threshold; for the i-th candidate access node in the set of k-th candidate access nodes, based on UAV data collected around the candidate access node... The set of sensor nodes within the range is used to determine the access benefit value of the candidate access node; based on the access benefit values ​​of each candidate access node in the k-th candidate access node set, a subset of candidate access nodes is selected from the k-th candidate access node set to obtain the k-th access node set; for the i-th access node in the k-th access node set, the hovering time of the access node is determined based on the set of sensor nodes within the data collection range of the UAV around the access node; based on the hovering time of each access node in the k-th access node set, a flight path is planned for the UAV from the entrance point to the exit point of the sub-region, passing through the nodes in the sub-region.

5. The UAV data collection method based on region division and path optimization according to claim 4, characterized in that, The method further includes: using the i-th and j-th access nodes in the k-th access node set as path nodes in the candidate flight path, determining the flight time and energy consumption required for the UAV to fly between the i-th and j-th access nodes, and thus determining the flight cost of the UAV flying between the i-th and j-th access nodes; and planning a flight path for the UAV from the entrance point to the exit point of the sub-region and passing through the nodes in the sub-region based on the hovering time of each access node in the k-th access node set, including: using the flight cost of the UAV flying between any two access nodes in the k-th access node set as the edge weight. Using each access node in the k-th access node set as a node, a k-th directed graph is constructed. Under the condition of satisfying the k-th constraint, with the objective of maximizing the total access revenue, it is determined whether each access node in the k-th access node set is accessed, so as to obtain the flight path of the sub-region. The k-th constraint includes: the sum of the total flight time and the total hovering time of the UAV in the k-th sub-region does not exceed a preset time; the sum of the total flight energy consumption and the total hovering energy consumption of the UAV in the k-th sub-region does not exceed a preset energy; the flight path of the UAV starts from the entrance point of the k-th sub-region and ends at the exit point of the k-th sub-region.

6. A UAV data collection method based on region division and path optimization according to claim 4, characterized in that, Based on the access benefit value of each candidate access node in the k-th candidate access node set, a subset of candidate access nodes is selected from the k-th candidate access node set to obtain the k-th access node set. This includes: selecting candidate access nodes whose access benefit value is greater than a preset threshold; or sorting each candidate access node in the k-th candidate access node set in descending order of access benefit value and selecting a subset of candidate access nodes based on the sorting result; or determining the access cost-effectiveness of each candidate access node in the k-th candidate access node set based on the access benefit value of each candidate access node and the total flight cost of the UAV flying from the candidate access node to other candidate access nodes, sorting each candidate access node in the k-th candidate access node set in descending order of access cost-effectiveness, and selecting a subset of candidate access nodes based on the sorting result.

7. The UAV data collection method based on region division and path optimization according to claim 1, characterized in that, Before controlling the UAV to perform a data collection task according to the global flight path, the method further includes: verifying whether the global flight path satisfies global constraints, the global constraints including at least one or more of the following: the total flight time of the UAV flying according to the global flight path does not exceed the maximum usable time of the UAV; the sum of the total flight energy consumption and the total hovering energy consumption of the UAV flying according to the global flight path does not exceed the maximum usable energy of the UAV; the sum of the total flight time and the total hovering time of the UAV flying according to the global flight path does not exceed the maximum execution time allowed for the data collection task; if it is detected that the global flight path does not satisfy the global constraints, perform at least one of the following adjustments until a global flight path that satisfies the global constraints is obtained: deleting at least one sub-region among the plurality of sub-regions; reducing the set of access nodes for at least one sub-region among the plurality of sub-regions; adjusting the access order of the plurality of sub-regions; controlling the UAV to perform a data collection task according to the global flight path, including: controlling the UAV to perform a data collection task according to a global flight path that satisfies the global constraints.

8. A method for collecting UAV data based on region division and path optimization according to any one of claims 1 to 7, characterized in that, The method further includes: during the data collection task performed by the UAV, based on the remaining usable time or remaining available energy of the UAV, performing at least one of the following adjustments: deleting at least one unvisited sub-region among the plurality of sub-regions; reducing the set of access nodes for at least one unvisited sub-region among the plurality of sub-regions; and adjusting the access order of each unvisited sub-region among the plurality of sub-regions.

9. A method for collecting UAV data based on region division and path optimization according to any one of claims 1 to 7, characterized in that, The method further includes: when the UAV arrives at the entrance point of any of the multiple sub-regions, re-acquiring the access revenue value of each access node in the access node set of the sub-region, and determining the change in the access revenue value of each access node in the sub-region. If the change in the access revenue value of at least one access node in the sub-region exceeds a preset change amount, without changing the entrance point and exit point of the sub-region, the method re-plans the flight path for the UAV from the entrance point to the exit point of the sub-region and passing through the nodes in the sub-region according to the re-acquiring access revenue value.

10. A UAV data collection system based on region division and path optimization, characterized in that, The system includes: a matrix determination module, used to determine the communication reachability matrix of the wireless sensor network based on the spatial location of each node in the wireless sensor network, wherein the elements in the communication reachability matrix are... The value of represents whether the i-th node and the j-th node in the wireless sensor network can communicate directly; the network topology determination module is used to determine the network topology diagram of the wireless sensor network based on the communication reachability matrix of the wireless sensor network, wherein the nodes in the network topology diagram are the nodes in the wireless sensor network, and the edges in the network topology diagram are determined based on the communication reachability matrix; the sub-region division module is used to divide the wireless sensor network into regions based on the network topology diagram of the wireless sensor network, with the objective that the spatial distance between any two nodes in the same sub-region is less than a target threshold and there is a connection relationship between adjacent sub-regions. The system is divided into multiple sub-regions; an entrance / exit determination module is used to determine the entrance and exit points of each sub-region based on the connection relationships between adjacent sub-regions; a flight path planning module is used to plan a flight path for the UAV from the entrance point to the exit point of each sub-region, passing through the nodes within that sub-region; and a global flight path planning module is used to plan a global flight path for the UAV within the wireless sensor network based on the flight paths of each sub-region and the connection relationships between adjacent sub-regions, and to control the UAV to perform data collection tasks according to the global flight path.

Images