Aircraft stall changing method based on near-end strategy optimization algorithm

The invention belongs to the technical field of aviation, and relates to an aircraft stall changing method based on a near-end strategy optimization algorithm. The method comprises the following steps: operating a flight simulator, collecting different trajectories of an aircraft, performing control strategy learning by using the flight trajectories of the aircraft based on an improved near-end strategy optimized intelligent agent, and transmitting state parameters of the aircraft to the intelligent agent through a UDP (User Datagram Protocol); a control strategy of the intelligent agent is transmitted to the flight simulator through a UDP, the ratio of beta distribution parameters is calculated by adopting beta distribution, so that assistance is provided for training planning, a smooth aircraft stall reward function is designed to improve the system training efficiency, and finally, the self-adaptive learning rate is combined, the sample efficiency of the flight path is fully utilized, and the system traini