Research On Automatic Landing Algorithm Of Carrier-based Aircraft And Design And Implementation Of Auxiliary Operation Control Training System

Under the influence of deck movement,narrow runway and complex sea condition,carrier-based aircraft landing has always been considered a high-risk task,so it is called "dance on the tip of the knife ",which is very difficult.To master this flight skill requires long-term training.In the existing pilot simulation training system,there are problems such as single function and lack of intelligence,lack of excellent teachers,difficulty in inheriting teachers’ experience and low training efficiency.Improving the efficiency and success rate of pilot training in the carrier-based aviation field is of great significance for cost reduction.During recent years,many scholars have applied deep reinforcement learning to autopilot,among which,the depth deterministic strategy gradient method(Deep Deterministic Policy Gradient,DDPG)has the advantages of fast convergence and excellent performance in continuous action,so it is suitable for decision research of autopilot system.A feasible way to solve the above problem is to train and generate the intelligent body with the ability to control the landing of shipborne aircraft by using the DDPG algorithm and to apply it to the training of shipborne aircraft pilots.The main achievements include:1)Automatic landing algorithm based on DDPG algorithm.On the basis of analyzing the influence of carrier movement and wind force on carrier-based aircraft landing,a Markov decision process(Markov Decision Process,MDP)model which can adapt to carrier movement and wind force is established.Based on the X-Plane flight simulation environment,the algorithm can control the automatic landing of the general aircraft model under various environmental conditions.2)Auxiliary operation control training system.The system can realize the generation,storage and management of the electronic instructor’s intelligent body for a variety of flight training subjects,and can use the relevant intelligent body to carry out flight assisted control training,and provide real-time flight control guidance.At the end of each training,the data of the training process can be compared and evaluated,and the evaluation results can be visualized to facilitate the students to review and review the training process,memorize and master the main points of operation.Finally,the effectiveness of the pilot intelligent auxiliary operation control training system based on the intelligent body control algorithm is carried out in this paper.The experiment shows that the control training carried out by the intelligent body can effectively improve the efficiency of pilot training.