Title :
Neural net-based adaptive linear quadratic control
Author_Institution :
Dept. of Autom. Control Eng., Feng Chia Univ., Taichung, Taiwan
Abstract :
A new indirect adaptive control scheme based on recurrent neural networks is proposed. The certainty equivalence principle is used to combine an adaptive law with a control structure derived from the linear quadratic (LQ) control problem. The proposed approach includes two sets of neural networks each with two feedback connected layers to solve for two types of algebraic matrix Riccati equations. One is for the Kalman filter and the other one is for the LQ controller design. The gradient algorithm is used as an adaptive law for identifying plant parameters and is used as the update rule for neural networks
Keywords :
Kalman filters; Riccati equations; adaptive control; control system synthesis; feedback; linear quadratic control; matrix algebra; neurocontrollers; recurrent neural nets; Kalman filter; algebraic matrix Riccati equations; certainty equivalence principle; feedback connected layers; gradient algorithm; indirect adaptive control; neural net-based adaptive linear quadratic control; recurrent neural networks; update rule; Adaptive control; Application software; Automatic control; Computer networks; Matrices; Neural networks; Programmable control; Recurrent neural networks; Riccati equations; Stability;
Conference_Titel :
Intelligent Control, 1997. Proceedings of the 1997 IEEE International Symposium on
Conference_Location :
Istanbul
Print_ISBN :
0-7803-4116-3
DOI :
10.1109/ISIC.1997.626450
