摘要： 采用了基于合同网的分布式规划方法，研究了战场环境中多无人机动态任务调度问题，并建立了数学模型，提出了分布式的任务调度体系结构，设计了一种基于代价变换的概率路标图路径规划算法，该算法能够在任务调度阶段，快速预估无人机执行不同任务的飞行航路，扩展了合同网协议，可在一次拍卖中“并发”进行多次交易，提高了任务调度的效率。通过多种合同类型的综合，解决了复杂战场态势下的任务调度问题。

关键词: 无人机, 任务调度, 合同网, 概率路标图

Abstract: This paper studies distributed planning techniques based on contract net for dynamic tasks scheduling of multiple unmanned aerial vehicles (UAVs) in battlefield, presents a multi-UAV tasks scheduling model, develops and a distributed tasks scheduling architecture is developed. A novel path planning algorithm based on probabilistic road maps is designed, which can plan predictive routes between the UAV and multiple tasks rapidly. Contract net protocols are expanded to simultaneity deal with multiple trades in one auction, so that negotiation efficiency is improved. By integreting several contract types which can work together, tasks scheduling in complex battlefield situations is solved.

Key words: unmanned aerial vehicle, task scheduling, contract net, probabilistic road maps

中图分类号: 

• TP242.6