基于MMDP的无人作战飞机任务分配模型研究  

RESEARCH ON TASK ALLOCATION MODEL FOR UCAV BASED ON MMDP

作　　者：李月娟[1] 吕永健[1] 常迁臻[2] 朱李云 

机构地区：[1]空军工程大学航空航天工程学院,陕西西安710038 [2]空军工程大学训练部,陕西西安710051 [3]空军驻郑州地区军事代表室,河南郑州450006

出　　处：《计算机应用与软件》2013年第7期276-279,286,共5页Computer Applications and Software

摘　　要：针对多无人作战飞机动态任务分配问题,以马尔科夫决策过程理论为基础,建立基于多智能体马尔科夫决策过程MMDP(Multi-Agent Markov Decision Process)的动态任务分配模型;系统中状态信息、目标的出现和收益值等均服从概率分布;改进的MM-DP寻优算法以多UCAV在有限时间内执行任务收益值最大为评价函数,并应用遗传算法在所有可选决策集中确定最优任务分配策略;数值仿真验证了算法的有效性。On the basis of Markov decision process theory, a dynamic task allocation model based on multi-agent Markov decision process （MMDP） is built aiming at the dynamic task allocation problem of multiple unmanned combat aerial vehicle （UCAV）. Status messages, target appearance and profit value in the system are all subjected to the probability distribution. The improved MMDP optimisation algorithm claims the maximum of the gains of the multi-UCAV during finite period as the evaluation function, and applies genetic algorithm to determine the best task assignment policy in all optional decision sets. Numerical simulation verifies the effectiveness of the algorithm.

关 键 词：多智能体 马尔科夫决策过程 无人作战飞机 动态任务分配 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术] V279[航空宇航科学与技术—飞行器设计]