Abstract:
Traditional designing methods lead to the big expenses by creation of new flight control systems. It is especially true in the presence of cross-communication and nonlinear connections, such as at helicopters. Therefore such methods, as a method of linearity a feedback or a method of dynamic inversion which cannot give improvement of characteristics because they depend from knowledge of system dynamics are developed. Adaptive neural network-based control can compensate unknown nonlinearity at of feedback linearization.
Keywords:control system, neural computer, adaptive methods, approximation of nonlinear connections.