Title :
Control of perturbed systems using neural networks
Author_Institution :
Dept. of Autom. Control Eng., Feng Chia Univ., Taichung, Taiwan
fDate :
9/1/1998 12:00:00 AM
Abstract :
Stability conditions for a perturbed plant control by a conventional robust controller and a neurocontroller are presented. The neural net-based direct inverse controller is proposed to aid the robust controller to further suppress the output error resulting from the unmodeled residuals. A procedure for determining the permissible network´s output under which the overall closed-loop system will be robustly stable is provided
Keywords :
closed loop systems; neurocontrollers; perturbation techniques; robust control; stability criteria; closed-loop system; direct inverse controller; neural networks; neurocontroller; output error suppression; perturbed plant control; perturbed system control; robust controller; stability conditions; unmodeled residuals; Control systems; Error correction; Flexible structures; Neural networks; Riccati equations; Robust control; Robust stability; Robustness; Uncertainty; Upper bound;
Journal_Title :
Neural Networks, IEEE Transactions on
