Title: Model-based learning control with non-repetitive initial conditions

Authors: Orest V. Iftime, Michel Verhaegen

Addresses: University of Groningen, Department of Econometrics, PO Box 800, 9700 AV Groningen, The Netherlands. ' Technical University of Delft, Delft Center for Systems and Control, Mekelweg 2, 2628 CD, Delft, The Netherlands

Abstract: In this paper we will focus on model-based learning control for discrete, time-varying, linear systems with non-repetitive initial state. We assume that the disturbances are repetitive in trial-domain. A particular form for the learning input update law is considered and the stability of a general learning input law is analysed. The refinement of the inputs design based on the learning concepts with non-repetitive initial state has the potential to enhance the performances of tracking control. Mechanical systems, chemical and other manufacturing processes such as batch distillation, heat treatment processes for metallic or ceramic products can benefit from the analysis performed in this paper.

Keywords: learning control; stability; non-repetitive initial state; automatic learning; real time.

DOI: 10.1504/IJISTA.2007.012480

International Journal of Intelligent Systems Technologies and Applications, 2007 Vol.2 No.2/3, pp.161 - 173

Published online: 19 Feb 2007 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article