A Novel Automatic Carrier Landing System with ADRC-Based Controller and LSTM-Based Compensation Method

Aiming at the problem of carrier-based aircraft automatic carrier landing, this paper proposes a new automatic carrier landing system (ACLS) design with controller based on active disturbance rejection control (ADRC) and deck motion compensation (DMC) method with long short-term memory (LSTM) prediction. First, the six degrees of freedom nonlinear dynamic model of carrier-based aircraft and the deck motion model of aircraft carrier are established. Second, to improve the anti-disturbance performance of ACLS, the ADRC-based controller is introduced to be the inner loop controller. Third, due to the advantage in predicting long time series data, LSTM recurrent neural network is used to make prediction of deck motion and combined with the controllers of ACLS, the method of DMC is introduced. Finally, the methods proposed in this paper are verified by simulation. The simulation results show that the ADRC-based controller have better anti-disturbance performance in windy environment compared with traditional controller, and combined with DMC method, the whole system can track the motion of desired touchdown point (DTP) with the tracking error is in range of ±0.6 m.