基于时延Petri网的多UUV系统的任务分配策略

针对多水下机器人(unmanned underwater vehicle,UUV)系统动态任务分配问题展开研究,在对系统的体系结构和任务分配机制进行分析的基础上,利用时延 Petri 网理论对系统的任务分配进行建模,提出了一种新的任务分配策略:群体协调层采用集中式任务分配策略,由主 UUV 将任务实时下达给各从 UUV;监控规划层采用基于适应度的分布式任务分配方式,充分发挥各异构 UUV 自身的智能。仿真结果验证了模型的合理性和任务分配策略的有效性。

The problem of dynamic task allocation in multiple unmanned underwater vehicles(UUVs)is stud- ied in this paper.Using timed Petri net theory to modeling for system task allocation based on the analysis of the system structure and task allocation mechanism,and a new task allocation strategy is presented:u- sing centralized task allocation strategy in colony cooperation layer and let leader UUV real timed to allocate task to each follower UUV;using distributed task allocation strategy in monitor plan layer based on fitness, fully bring the intelligence of each different construction UUV into play.The simulation result has validated the rationality of the model and the validity of the task allocation strategy.

国家863高科技发展计划项目(2006AA04Z262)资助