基于集散式模型预测控制的多无人机协同分区搜索

针对多无人机在对大范围目标区域执行协同搜索任务时搜索资源分配不均、容易因频繁转场造成资源浪费等问题,借鉴集中式控制和分布式控制结构的优点,建立了集散式多无人机(unmanned aerial vehicle,UAV)协同搜索结构体系,通过聚类分析和V图划分等方法对目标区域进行分区,结合各子区域任务特点对无人机群进行搜索任务分配,并采用一种经改进后可有效增大UAV预测范围的预测控制模型,研究了动态环境下多UAV集散式协同分区搜索问题,最后,将所提方法与常见几种协同搜索方法进行对比仿真,获取仿真结果验证了所提方法在目标发现概率和搜索效率方面的有效性和优越性.

Multi-unmanned aerial vehicles cooperative search based on central-distributed model predictive control

When multiple unmanned aerial vehicles (UAVs) implement their surveillance and search operations in a broad-expanse uncertain environment, there exists resource-wasting on repetitive searching or frequent transition from one cell to another based on some traditional search algorithms. Promoted by the advantages in the central and distributed control architectures, we propose a multi-UAV cooperative search strategy to partition the target area by clustering analysis and voronoi partitioning. An improved predictive model is presented to design the multi-UAVs search strategy under dynamic environment. Finally, comparing various simulation results demonstrate the validity of the proposed cooperative search strategy.