Application of RBF Network in System Identification for Flight Control Systems

The flight control system (FCS) is a complex nonlinear multi-input and multi-output) system, it is very difficult to identify the model of this system. The nonlinear relation of I/O data can be expressed by artificial neural network (ANN). The ANN can fit the any function accurately by studying. In this paper, the FCS of a type of fighter is identified by radial basis function (RBF) network. The training algorithm of the RBF network is improved by a grouping optimizing method and a new training algorithm. Simulation results about application of this network with new algorithm were given. The results show that the complex FCS can be identified by artificial neural network accurately.