Design for Adaptive Flight/Propulsion Control System Based on Wavelet Neural Networks

陈谋,姜长生,吴庆宪,曹邦武
DOI: https://doi.org/10.3969/j.issn.1000-8055.2003.03.022
2003-01-01
Abstract:In this paper, an adaptive H∞ control method based on wavelet neural network is proposed for the nonlinear uncertain systems. The controller is made up of equivalent controller and H∞ controller. The wavelet neural networks are used to approximate the nonlinear functions and the approximation errors of the neural networks are introduced to the adaptive law in order to improve the quality of the system. H∞controller is designed for attenuating the external disturbance and approximation errors of the neural networks. The controller can not only guarantee the stability of the overall system,but also attenuate the effect of the external disturbance and approximation errors of the neural networks to a prescribed level. At last, based on forgoing designed control method a flight/propulsion control system has been designed and high angle attack flight simulation has been made. The simulation results suggest that the flight/propulsion control system is valid. They validate the availability of this control method for the uncertain nonlinear system too.
What problem does this paper attempt to address?