有人/无人作战智能体分布式任务分配方法

Distributed task allocation method of manned/unmanned combat Agents

有人/无人作战智能体任务联盟是面向分布式网络化作战体系提出的一种作战样式,任务分配问题是研究任务联盟指挥策略的关键点之一。以有人/无人作战智能体任务联盟为研究对象,提出与之适应的分布式体系结构。将任务执行质量引入任务分配问题建模中,以拍卖算法为基础,通过编组整体拍卖、方案预处理等改进机制,有效降低了个体方案的构建开销,并在约束的时间内实现任务联盟的动态任务分配。针对作战想定进行了仿真计算,结果表明算法能在有限拍卖次数下给出接近理想优化效果的分配方案。

The manned/unmanned combat Agents task coalition is a new combat mode, which is geared to the needs of the distributed networked operations system. Task allocation is one of the key points in studying the task coalition^s command policy. The distributed system architecture adapted to the task coalition is pro- posed. The task implement quality is introduced into task allocation modeling. Through the improved mecha- nism of grouping auction as a whole and scheme predisposition, the formed overhead of a unity scheme is re- duced. The method rooted in the auction algorithm can realize the dynamic task allocation of the task coalition within the restraint time. Simulation results based on combat scenarios indicate that the algorithm can present the allocation scheme closed to ideal optimal effect in a limited auction cycle.