Journal of Automatic Control and Computer Sciences, vol.38, no.3, pp.53-63, 2004 (SCI-Expanded)
In recent years, aircraft accidents have especially increased on Concorde aircrafts. Due to out of control of the aircrafts, many people have lost their lives. Neural network controllers can be accepted as an alternative to control this kind of planes. In this study, Neural network control system was employed to control nose’s angle of supersonic type Concorde aircrafts. The designed Model Reference Adaptive Neural Controller (MRANC) is feedforward multilayered perceptron structure. Backpropagation (BP) algorithm was utilized to update weights of the neural controller. For comparison, standard PID control system was also used with empirically selected gain parameters. Consequently, between two approaches, Neural networks have superior performance to control such aircrafts.