摘要
针对目前无人机集群对多目标进行协同侦察时,易重复侦察目标,进而导致侦察效率低的问题,提出了一种多无人机多目标协同侦察航迹规划算法。首先,优化了K-means聚类算法的评价标准,使目标集合的聚类结果更加稳定,同时也降低了目标被重复侦察的概率。然后,利用改进的离散粒子群算法求解侦察序列,来降低整体任务的时间代价。最后依据侦察序列生成各无人机任务航迹。仿真结果表明,该算法不仅能够有效避免目标被重复侦察,而且相较于基因算法和标准离散粒子群算法,在4 架无人机观测 30个目标的仿真条件下,将时间代价降低24%,其收敛速度较快,求解精度更高。
Aiming at the problem of low reconnaissance efficiency caused by repeatedly reconnoitring the target when multiple unmanned aerial vehicles (multi-UAVs) perform cooperative reconnaissance on multiple targets, a path planning algorithm for multi-UAVs cooperative reconnaissance multi-target is proposed. Firstly, the evaluation criteria of K-means clustering algorithm is optimized, which makes the clustering results of the target set more stable and also reduces the probability that the target is repeatedly reconnoitred. Then the improved discrete particle swarm optimization algorithm is used to solve the reconnaissance sequence, which reduces the time cost of the entire mission. Finally, the mission track of each UAV is generated based on the reconnaissance sequence. Simulation results show that the proposed algorithm can not only effectively avoid repeated reconnaissance on a target, but also has faster convergence speed and higher solution accuracy. Compared with the genetic algorithm and the standard discrete particle swarm optimization algorithm, the proposed algorithm reduces the time cost by 24% under the simulation condition that 4 UAVs observe 30 targets.
作者
庞强伟
胡永江
李文广
PANG Qiangwei;HU Yongjiang;LI Wenguang(Department of Unmanned Aerial Vehicle Engineering, Army Engineering University, Shijiazhuang 050003, China)
出处
《中国惯性技术学报》
EI
CSCD
北大核心
2019年第3期340-348,共9页
Journal of Chinese Inertial Technology
基金
国家自然科学基金(51307183)
军内科研(ZS2015070132A12007)
关键词
多无人机
多目标
协同
聚类
离散粒子群算法
multiple unmanned aerial vehicles
multi-target
cooperative
clustering
discrete particle swarm optimization
