CN-120742912-B - Large-scale unmanned mechanism type conversion method and device based on descending order strategy

CN120742912BCN 120742912 BCN120742912 BCN 120742912BCN-120742912-B

Abstract

The invention discloses a large-scale unmanned mechanism type conversion method and device based on a descending strategy, wherein the method comprises the steps of arranging unmanned aerial vehicle clusters in an original configuration in descending order based on the node distance between each initial node in the original configuration and a reference point; similarly, the unmanned aerial vehicle clusters in the target configuration are also arranged in a descending order, all initial nodes in the obtained original configuration sequence and all target nodes in the target configuration sequence are sequentially subjected to global pairing to generate a plurality of pairing, the pairing is arranged in a descending order according to an ideal path, pairing optimization processing is carried out on a preselected pairing sequence obtained by the descending order to obtain a quasi-pairing sequence, path obstacle avoidance processing is carried out on each pairing in the alignment pairing sequence, and path planning of the unmanned aerial vehicle configuration is generated. Therefore, the calculation complexity can be reduced from O (N|) to O (N 2 +δM 2 +L multiplied by N), M and L < < N, and unmanned aerial vehicle energy consumption balance can be realized on the basis of ensuring the optimal path, so that the management efficiency of the large-scale unmanned aerial vehicle cluster is improved.

Inventors

LAN BIN
WANG LONGFEI
WEI BANGGUO
CHEN XIANGYANG
LIU JINMING
WANG HAO

Assignees

江淮前沿技术协同创新中心

Dates

Publication Date: 20260512
Application Date: 20250604

Claims (10)

1. A method for large-scale unmanned mechatronic transformation based on a descending strategy, comprising: based on the node distance between each initial node in the original configuration and the reference point, the unmanned aerial vehicle clusters in the original configuration are arranged in a descending order to generate an original configuration sequence; The unmanned aerial vehicle cluster in the target configuration is arranged in a descending order based on the node distance between each target node in the target configuration corresponding to the original configuration and the reference point, and a target configuration sequence is generated, wherein the starting node is used for indicating the azimuth information of the unmanned aerial vehicle in the original configuration; sequentially performing global pairing on all the initial nodes in the original configuration sequence and all the target nodes in the target configuration sequence to generate a plurality of pairing; arranging the pairs in descending order according to ideal paths corresponding to the pairs, and carrying out pairing optimization treatment on a preselected pairing sequence obtained by descending order to obtain a quasi-pairing sequence; Performing path obstacle avoidance processing on each pair in the quasi-pairing sequence to generate a path plan of an unmanned aerial vehicle configuration; The method comprises the steps of arranging the pairs in descending order according to ideal paths corresponding to the pairs, carrying out pairing optimization processing on a preselected pairing sequence obtained by descending order to obtain a quasi-pairing sequence, determining ideal paths corresponding to each pair in the pairs, sequencing the pairs according to descending order of the ideal paths to obtain a preselected pairing sequence, selecting the first N pairs from the preselected pairing sequence in a node random sampling mode, carrying out local optimization processing on subsequences formed by the first N pairs to generate an optimized pairing sequence, and carrying out monomer optimization processing on the optimized pairing sequence to obtain the quasi-pairing sequence.
2. The method of claim 1, wherein the sequentially globally pairing all the initial nodes in the original configuration sequence and all the target nodes in the target configuration sequence to generate a plurality of pairs comprises: Selecting a first initial node from the original configuration sequence, and selecting a first target node from the target configuration sequence; if the node distance corresponding to the first starting node is greater than the node distance corresponding to the first target node, determining the path length between the first starting node and each target node in the target configuration sequence, and selecting the target node with the shortest path length from the target configuration as the pairing node of the first starting node; If the node distance corresponding to the first initial node is smaller than the node distance corresponding to the first target node, determining the path length between the first target node and each initial node in the original configuration sequence, and selecting the initial node with the shortest path length from the original configuration as a pairing node of the first target node; Removing pairing nodes successfully paired from the original configuration sequence and the target configuration sequence respectively, and generating a first updating sequence corresponding to the original configuration sequence and a second updating sequence corresponding to the target configuration sequence; and repeating the steps in sequence until all unmanned aerial vehicles in the original configuration sequence or the target configuration sequence finish pairing, and generating a plurality of pairing.
3. The method of claim 1, wherein the selecting the first N pairs from the pre-selected pair sequence in a node random sampling manner, and performing local optimization on the subsequences formed by the first N pairs to generate an optimized pair sequence, comprises: dividing the preselected pairing sequence into two parts according to the random sampling parameter N to obtain a first subsequence formed by the first N pairing and a second subsequence formed by the rest pairing; based on the principle of shortest total ideal paths in the first subsequence, performing pairing optimization on all unmanned aerial vehicles in the first subsequence according to a Hungary algorithm to obtain an updated subsequence; determining ideal paths of each pairing in the updated subsequence, and adding the updated subsequence to the second subsequence according to an ideal path descending order arrangement mode to generate a candidate pairing sequence; Determining a random sampling parameter N, selecting a third subsequence formed by the first N pairs from the candidate pair sequences, and determining an average path corresponding to the third subsequence based on ideal paths corresponding to each pair in the third subsequence; and taking the candidate pairing sequence as a new preselected pairing sequence, and sequentially repeating the steps until the average path of the third subsequence is not changed any more, ending the operation, and generating an optimized pairing sequence.
4. The method of claim 1, wherein the performing monomer optimization on the optimized pairing sequence to obtain a quasi-pairing sequence comprises: Based on the ideal path corresponding to each pairing in the optimized pairing sequence, descending order arrangement is carried out on the optimized pairing sequence, and an ordered pairing sequence is generated; selecting the pairing with the largest ideal path from the ordered pairing sequence as a target pairing; Executing unmanned aerial vehicle node replacement operation aiming at the target pairing, and generating updated target pairing; if the ideal path of the updated target pairing is smaller than the original target pairing, replacing the original target pairing by the updated target pairing; And sequentially repeating the steps until the total length of the ideal path corresponding to the optimized pairing sequence is kept unchanged, ending the unmanned aerial vehicle node replacement operation in the target pairing, and outputting a quasi-pairing sequence.
5. The method of claim 1, wherein the performing path obstacle avoidance processing on each pair in the quasi-pairing sequence to generate a path plan of an unmanned aerial vehicle configuration comprises: numbering unmanned aerial vehicles of the initial node corresponding to each pairing in the quasi-pairing sequence to obtain a plurality of numbered unmanned aerial vehicles; the method comprises the steps of determining a current numbered unmanned aerial vehicle corresponding to the current pairing according to any current pairing in the quasi-pairing sequence, controlling the current numbered unmanned aerial vehicle to receive path plans sent by other numbered unmanned aerial vehicles in real time, broadcasting the path plans of the current numbered unmanned aerial vehicles in real time, and acquiring a current path node and a target node corresponding to the current pairing, wherein the current path node is used for indicating a starting node or an intermediate node between the starting node and the target node; Planning a next path node adjacent to the current path node according to a static obstacle avoidance processing method based on the current path node and the target node, and obtaining a first path plan of the next path node; If the received path plan represents that a dynamic obstacle exists between the current path node and the next path node and the dynamic obstacle is not a path conflict, determining that the first path plan is not established, and re-planning the next path node according to a dynamic obstacle avoidance processing method to generate a second path plan corresponding to the next path node; if the received path planning characterizes that a dynamic obstacle exists between the current path node and the next path node and the dynamic obstacle is a path conflict, re-planning the next path node based on a dynamic planning method to generate a second path planning corresponding to the next path node; Taking the next path node corresponding to the second path plan as a quasi path node; If the received path plan represents that no dynamic barrier exists between the current path node and the next path node, the next path node corresponding to the first path plan is taken as a quasi path node; Sequentially repeating the steps until the unmanned aerial vehicle with the current number reaches a target node through different quasi-path nodes, and generating a path plan corresponding to the current pairing by taking the reached target node as a fixed barrier; and generating a path plan of the unmanned aerial vehicle based on the path plan corresponding to each pairing in the quasi-pairing sequence.
6. The method of claim 5, wherein the planning, based on the current path node and the target node, of a next path node adjacent to the current path node according to a static obstacle avoidance process to obtain a first path plan for the next path node, comprises: acquiring attractive force of the target node to the current path node and attractive force potential field corresponding to the attractive force; Acquiring repulsive force of a static obstacle to the current path node and a repulsive force potential field corresponding to the repulsive force; determining a direction corresponding to a next path node adjacent to the current path node based on the attractive force and the repulsive force; determining position information corresponding to a next path node adjacent to the current path node based on the attractive potential field and the repulsive potential field; and determining the position information and the direction corresponding to the next path node as the azimuth information of the next path node, thereby obtaining a first path plan of the next path node.
7. The method of claim 1, wherein the step of determining the position of the substrate comprises, Determining a centroid of the target configuration based on the positional information of each drone in the target configuration; determining a node distance between each starting node in the original configuration and the centroid; Based on the node distance corresponding to the initial node, the unmanned aerial vehicle clusters in the original configuration are arranged in a descending order to generate an original configuration sequence; determining a node distance between each target node in the target configuration and the centroid; and based on the node distance corresponding to the target node, arranging unmanned aerial vehicle clusters in the target configuration in a descending order to generate a target configuration sequence.
8. The method as recited in claim 1, further comprising: extracting outer surface contour points from a three-dimensional model corresponding to the first unmanned aerial vehicle configuration, and performing scaling treatment on the extracted outer surface contour points to generate an original configuration; Acquiring a second unmanned aerial vehicle configuration corresponding to the first unmanned aerial vehicle configuration; and extracting the outline points of the outer surface from the three-dimensional model corresponding to the second unmanned aerial vehicle configuration, and scaling the extracted outline points of the outer surface to generate a target configuration.
9. A large-scale unmanned aerial vehicle configuration conversion device based on a descending order strategy, comprising: The first generation module is used for carrying out descending order arrangement on unmanned aerial vehicle clusters in the original configuration based on the node distance between each initial node in the original configuration and the reference point, and generating an original configuration sequence; The system comprises a first generating module, a second generating module, a target configuration sequence, a target node and a target node, wherein the first generating module is used for generating a target configuration sequence by arranging unmanned aerial vehicle clusters in the target configuration in a descending order based on the node distance between each target node in the target configuration corresponding to the original configuration and a reference point; The global pairing module is used for sequentially carrying out global pairing on all the initial nodes in the original configuration sequence and all the target nodes in the target configuration sequence to generate a plurality of pairs; the optimization processing module is used for sequencing the pairs in a descending order according to ideal paths corresponding to the pairs and carrying out pairing optimization processing on the obtained preselected pairing sequence to obtain a quasi-pairing sequence, and comprises a determining unit, a local optimization unit, a monomer optimization unit, a single optimization unit and a quasi-pairing sequence, wherein the determining unit is used for determining the ideal paths corresponding to each pair in the pairs and sequencing the pairs according to the descending order of the ideal paths to obtain the preselected pairing sequence; And the path obstacle avoidance processing module is used for carrying out path obstacle avoidance processing on each pairing in the quasi-pairing sequence and generating a path plan of the unmanned aerial vehicle.
10. A computer readable medium having stored thereon a computer program which, when executed by a processor, implements the method of any of claims 1 to 8.

Description

Large-scale unmanned mechanism type conversion method and device based on descending order strategy Technical Field The invention belongs to the technical field of unmanned aerial vehicles, and particularly relates to a large-scale unmanned mechanism type conversion method and device based on a descending strategy. Background The rapid development of unmanned aerial vehicle technology makes unmanned aerial vehicle have extensive application in civil fields such as military field, video taking photo by plane, environmental monitoring, geographical mapping, electric power line patrol, environmental monitoring, etc. In the face of increasingly complex task demands, a single unmanned aerial vehicle is difficult to independently complete tasks, so that a mode that multiple unmanned aerial vehicles cooperatively execute tasks is carried out. Unmanned aerial vehicle path planning is the core of unmanned aerial vehicle cooperative control, so research many unmanned aerial vehicle cooperative path planning has very important meaning. The multi-unmanned aerial vehicle collaborative path planning essence is to coordinate paths among a plurality of unmanned aerial vehicles to complete collaborative tasks. The existing path method based on graph search and sampling has high efficiency and excellent global optimization performance in environment search, however, the algorithm has the problems of dynamic obstacle avoidance and communication in a large-scale environment. Path planning methods based on heuristic algorithms, such as ant colony optimization methods, particle swarm optimization methods, bee colony optimization methods, genetic algorithms and simulated annealing algorithms, can effectively plan paths and distribute tasks in complex, dynamic and multi-constraint environments, but have the limitations of high calculation cost, easy trapping in local optima and the like. The learning algorithm can adjust the strategy by continuously learning new information in the environment, has strong adaptability, is suitable for dynamic and uncertain environments, but has strong dependence on large-scale and high-quality training data, and can greatly reduce the performance of the algorithm when the data is insufficient or inaccurate, and has poor algorithm interpretability, thereby being unfavorable for the debugging and optimization of the system. The existing multi-unmanned aerial vehicle collaborative path planning algorithm is mainly oriented to the problems of dynamic obstacle avoidance and search coverage, and has no excessive relation to the field of multi-unmanned aerial vehicle mechanism type conversion. The computational complexity of the existing algorithm grows exponentially along with the increase of the cluster size, and the global task planning is not suitable for large-scale unmanned aerial vehicle cluster configuration transformation. In addition, the general optimization method only considers factors such as shortest path and minimum energy of task allocation, and the path standard deviation factor is often ignored, so that unmanned aerial vehicle energy consumption is unbalanced, and the management difficulty of unmanned aerial vehicle clusters is increased. Disclosure of Invention The invention provides a large-scale unmanned aerial vehicle mechanism type conversion method and device based on a descending order strategy, which can reasonably plan paths among all pairs, so that the purposes of minimum path and balanced energy consumption of all pairs are achieved, and the management efficiency of a large-scale unmanned aerial vehicle cluster is further improved. According to a first aspect of the embodiment of the invention, a large-scale unmanned mechanism type transformation method based on a descending order strategy is provided, and the method comprises the steps of conducting descending order arrangement on unmanned aerial vehicle clusters in an original configuration based on node distances from each initial node to a reference point in the original configuration, generating an original configuration sequence, conducting descending order arrangement on unmanned aerial vehicle clusters in a target configuration based on node distances from each target node to the reference point in a target configuration corresponding to the original configuration, generating a target configuration sequence, wherein the initial nodes are used for indicating azimuth information of unmanned aerial vehicles in the original configuration, the target nodes are used for indicating azimuth information of unmanned aerial vehicles in the target configuration, conducting global pairing on all initial nodes in the original configuration sequence and all target nodes in the target configuration sequence in sequence, generating a plurality of pairs, conducting descending order arrangement on the pairs according to ideal paths corresponding to the pairs, conducting pairing optimization processing on the pairing sequences obtained t