Research on Direct Lift Landing Control Based on Neural Network

Aiming at the optimization of control parameters in the landing control phase of carrier based aircraft, a neural network based direct lift landing control method for carrier based aircraft is proposed. Based on the original direct lift landing control structure, the landing control structure is divided into direct force loop, trajectory loop, angle of attack loop and speed loop. The direct force is introduced into the control system to control the trajectory angular rate of the carrier based aircraft, and then the altitude is controlled through the trajectory. The elevator is used to balance part of the pitching moment brought by the direct force, keep the angle of attack constant, and The accelerator is used to keep the speed constant. Keep the throttle constant. By introducing neural network into the direct force loop, the controller parameters are optimized online to improve the convergence speed. Finally, the interference of deck motion and carrier air wake as well as Monte Carlo pull deviation condition are added to the simulation experiment. The results show that the trajectory correction speed of the ship based aircraft direct lift landing control based on neural network is faster, the overshoot is smaller, and the accuracy is higher.