Planning for target system striking based on Markov decision process

Planning for targets selection and striking is an important part during the military decision process. In this paper, as to optimize the striking process against the military target system, in which there are relationships between different targets and each target having the capability to recover from the damage, the structure and the recovering mechanism of the target system is modeled. In order to generate a multi-phase striking planning which can destroy the target system efficiently, the Markov Decision Process based planning method is proposed, and a heuristic is calculated to reduce the search space of the problem. Efficiency of the model is showed by a case, which shows that commander can raise their decision efficiency by this method.