Nonlinear Dynamic Inverse Carrier Landing Control Based on the Minimum Parameter Learning Method of RBF Network

Considering the problems of dynamics coupling between six degrees of freedom, internal modeling error and external wake interference in the process of carrier landing, this paper uses RBF neural network with a novel NN parameter learning method called minimum parameter method to estimate and compensate the composite interference, so as to realize the decoupling control and robust control. Firstly, according to the principle of time scale separation, the multi-layer dynamic inverse control of 6DoF dynamics is conducted. Then, the RBF neural network is designed for dynamic compensation to solve the problem of compound interference. To meet the requirements of real-time control, the minimum parameter learning method is applied to transform the RBF multi-parameters estimation into a single parameter estimation which is proved feasible with the Lyapunov method. The simulation results show that the system has good robust performance to ensure the accuracy and safety of carrier landing.