A Hebbian feedback covariance learning paradigm for self-tuning optimal control

Author(s):  
D.L. Young ◽  
C.-S. Poon
1987 ◽  
Vol 31 ◽  
pp. 299-304
Author(s):  
Reynaldo R. Medina ◽  
Kenji Jinno ◽  
Toshihiko Ueda ◽  
Akira Kawamura

1989 ◽  
Author(s):  
Hung-Hsu Fred Chen ◽  
Dennis A. Guenther

1992 ◽  
Vol 114 (1) ◽  
pp. 139-144
Author(s):  
K. W. Junk ◽  
R. R. Fullmer ◽  
R. C. Brown

This paper investigates the temperature control of a two-bed fluidized combustor using a self-tuning control algorithm to vary secondary air flow rate. The controller consists of a recursive least-squares parameter estimator, an observer, and a linear optimal control design procedure. This combination enables the controller to estimate the system parameters and update the feedback gains when necessary. Further, this study addresses the tracking form of optimal control, accomplished by augmenting the state vector with an integrator. The self-tuning control algorithm was compared with a simple PI controller, which was tuned using the Ziegler-Nichols method. In this study, self-tuning control provided improved performance over classical control. Compared with conventional, constant-gain control, self-tuning control reduced steady-state variance by a factor of 6.67 while maintaining good tracking characteristics.


Sign in / Sign up

Export Citation Format

Share Document