Robust non-linear feedback control for BTT missile with NN-based uncertainty estimation

A robust non-linear feedback control strategy combined with neural network (NN) estimator is presented, for the non-linear model of bank-to-turn (BTT) missile with modelling uncertainties. The non-linearities in the missile dynamics are taken into account, including coupling items between three channels. Standard feedback linearisation is implemented to linearise and decouple the nominal system. In the presence of unknown bounded uncertainties, the performance will deteriorate because the precise model cannot be obtained. Then, linear matrix inequality (LMI) based guaranteed cost control (GCC) is adopted to solve the robust control problem for the linearised uncertain models. Further, adaptive neural network estimators, which use Lyapunov based tuning rules, are integrated in the control strategy to eliminate the effect of high-order uncertain terms. Simulation results on a specified BTT missile model are provided to demonstrate the feasibility and effectiveness of the proposed approach, which achieves more improved tracking performance comparing with conventional method.