基于混合蚁群算法的MAS任务分配

Task allocation for MAS based on hybrid ant colony algorithm

在MAS(多agent系统)中,由于任务的复杂性和agent求解问题能力的不同,任务和agent不再是传统的一对一的关系。为解决MAS的任务分配问题,提出了任务与agent之间多对多的任务分配模式。首先建立了任务分配的数学模型,并导出分配优化的目标函数;其次利用混合蚁群算法快速收敛和分布式求解的特点实现任务分配的组合优化。对实验仿真的结果分析表明,多对多的任务分配模式能够明显提升多agent系统的性能。

In multi-agent systems （ MAS）, the relation of tasks and agents are no longer the traditional one-to-one relation because of complexity of tasks and performance difference between respective agents. To solve task allocation problem in MAS, put forward a many-to-many task allocation mechanism. First, established the mathematical model of task allocation and further deduced the target function. Then applied a hybrid ant colony algorithm to accomplish combinatorial optimization of task allocation. Finally, the simulation experiments show that the many-to-many task allocation mechanism can significantly enhance the performance of MAS.