Designing stable iterative learning control systems from frequency based repetitive control designs
Conference proceedings article
Authors/Editors
Strategic Research Themes
No matching items found.
Publication Details
Author list: Panomruttanarug B., Longman R.W., Phan M.Q.
Publication year: 2012
Volume number: 142
Start page: 2893
End page: 2912
Number of pages: 20
ISBN: 9780877035770
ISSN: 0065-3438
eISSN: 0065-3438
Languages: English-Great Britain (EN-GB)
Abstract
Repetitive control (RC) and iterative learning control (ILC) design control systems that aim for zero tracking error in repeating situations. ILC applies to spacecraft problems that perform repeated scanning maneuvers with a fine pointing instrument. Very effective RC design methods use steady state frequency response approaches. They cannot be applied directly to ILC because ILC is a finite time problem, and asks for zero error not only in the part of the desired trajectory after transients have decayed, but also asks for zero error during the transients as well. This work converts these effective RC design methods so that they apply to the ILC problem, and produce convergence to zero tracking error that is monotonic with iterations in the sense of the Euclidean norm. One first fills the ILC gain matrix with the RC gains, and then adjusts a few gains associated with the first few time steps, using a steepest descent or other similar algorithm. We show that adjusting only one gain can be sufficient to produce asymptotic stability and monotonic convergence. The method is simple to apply, and allows one to make use of a very effective RC design method in ILC applications, and represents a powerful way to apply frequency response ideas to finite time ILC problems.
Keywords
No matching items found.
Contact Information