舰载机出动回收调度策略生成方法

为了提高航母舰载机的出动回收能力，从提高出动回收调度过程的自动化程度入手，采用了最初在机器人导航和控制领域提出的学徒学习理论来解决调度优化问题。通过建立基于马尔科夫决策过程的舰载机出动回收过程仿真模型，以专家的示范调度操作为学习目标，采用学徒学习理论中的乘法权重法构建出动回收调度方法，并根据舰载机在集中出动和连续出动这两种典型调度工况下的具体算例研究，将得到的结果与专家示范操作结果进行对比，认为该方法具有较好的优化效果和较高的实用性。 

A Method of Policy Automated Generation for Carrier Aircraft Sortie and Recovery Scheduling

To improve the capability of carrier aircrafts sortie and recovery, a theory of apprenticeship learning, came of the robot navigation and control domain, was applied to automate the process of sortie and recovery scheduling. Firstly, a simulation model of aircrafts sortie and recovery was established based on the frame of Markov decision process. Then, taking an expert''s demonstration operation as learning former, an optimized scheduling policy was created with the multiplicative weights apprenticeship learning algorithm. Compared with the optimization results of the two typical research cases in the condition of group sortie and continuous sortie with the expert''s demonstration, the algorithm shows a better performance and function.