The mission planning of the carrier-based aircraft dispatching and recovery operation on the aircraft carrier deck refers to formulating a reasonable and feasible action plan for a given flight mission. This course of action is highly relevant to the efficiency of the carrier's mission, the safety of deck operations, and the cost. The task planning problem of carrier-based aircraft dispatching and recovery operation on the aircraft carrier deck has the characteristics of diverse tasks, distinct layers and complex constraints, and it is difficult to solve. In the future, more and more unmanned aerial vehicles (UAVs) will be carried on the aircraft carrier. It is an inevitable trend that the automated planning method will play an important role in the field of aircraft carrier deck mission planning. The purpose of this paper is to study the automatic planning algorithm considering the complex constraints for the task planning problem of the carrier-based aircraft dispatching and recovery operation on the aircraft carrier deck. Firstly, this paper constructs the task planning problem model of the carrier-based aircraft dispatching and recovery operation in the background of the "Nimitz" class aircraft carrier. Then, according to the characteristics of the problem, a task planning algorithm for dispatching and recycling operations with complex constraints is proposed. The algorithm uses the idea of task decomposition in hierarchical task network planning to decompose tasks to improve planning efficiency, and describes the planning problem as a multi-stage decision-making problem.
question. Then, for the three decision-making links in the algorithm, a reinforcement learning algorithm is designed to learn the strategies of task lead time prediction, carrier aircraft selection and transfer position selection based on historical plans, so as to guide the planner to better determine the parameters and parameters of the decision-making link. Rules to improve the performance of action plans. Finally, through an example analysis...