Integrated guidance and control for homing missiles via neural network

Abstract

To obtain much smaller miss distance and even shorter response time, the problem of integrated guidance and control for homing missiles in the pitch plane is considered. As the relative motion equations and the missile dynamics equations are built in different coordinates, the integrated guidance and control model is difficult to build. Also, during the flight, the accurate full-envelope aerodynamic parameters of the missile is not available. In this paper, the proper and compact integrated model is formulated under some assumptions and further changed into a common form of cascaded nonlinear systems with unknown uncertainties. By adopting the backstepping control approach, the stability of the entire states is guaranteed. An structure of critic neural networks is designed to approximate the uncertainties brought by the variety of aerodynamic parameters. The numerical simulation results demonstrate the effect of the proposed design scheme.