This chapter presents a number of simple methods for accelerating the convergence of the Norm-Optimal Iterative Learning control (NOILC) algorithm as measured by the number of online experiments required to achieve a desired tracking accuracy. The approaches require algorithm modifications and/or additional offline, model-based calculations but have discernible beneficial effects although some reduction in robustness can be anticipated. The notation and models assumed are precisely those used in the NOILC Chap. 5. A familiarity with the ideas, techniques, and examples used in that chapter will be of great value to the reader.
Accelerating NOILC Convergence
Advances in Industrial Control
13.06.2025
40 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
Accelerating convergence of the Frank-Wolfe algorithm
Elsevier | 1984
|Accelerating Convergence of Iterative Image Restoration Algorithms
British Library Conference Proceedings | 2007
|