Robust dynamic inversion control for BTT missiles based on neural network

A nonlinear control scheme, which is a mixed scheme of inversion system method and neural network adaptation theory, is applied to the design of bank-to-turn (BTT) missiles in this paper. First, the control model, which is based on inversion system method and neural network adaptation theory, is built. Second, the progressive convergence of tracking error is proved theoretically. Then, the on-line neural network is used to dynamically eliminate the influence of inversion error on system; hence this mixed scheme overcomes the shortage of nonlinear dynamical inversion control scheme where a precise mathematic model is necessary. Finally the validity of the control scheme is certified by digital simulation.