多UCAV(U nm anned Com bat A ir V eh icle)任务分配是无人机作战决策的重要内容,是一类求解困难的组合优化问题,目前尚无成熟高效的解决方案。通过马尔可夫决策过程(M DP)将任务分配问题化解为分阶段序列决策过程,并对当前阶段决策状态采用单一整数编码,提出了基于有限阶段M DP的求解算法。通过仿真实验表明:该算法可以快速精确地解决多UCAV任务分配的问题。 Multi-UCAV mission assignment is an important part of UAV combat decision making. It is a kind of combinatorial optimization problem which is difficult to solve. There is no mature and efficient solution. Through the Markov Decision Process (MDP), the task allocation problem is decomposed into a phased sequence decision process and a single integer encoding is used for the decision phase at the current stage. A solution algorithm based on the finite phase M DP is proposed. The simulation results show that this algorithm can quickly and accurately solve the problem of multi-UCAV assignment.