A New Arimoto-Type Algorithm to Estimate States for Repetitive Processes: Iterative Learning Observer (ILO)

Iterative learning control (ILC) has been established as a very powerful technique to achieve high performance control for repetitive processes. In addition to the ILC problem, in the literature researchers have considered related paradigms such as iterative feedback tuning and iterative parameter estimation and identification. In the paper we introduce another related problem: the iterative learning observer. This dual problem establishes algorithms that can achieve high performance estimation of states for iterative processes. In order to solve the dual problem, this paper develops an estimation algorithm that estimates states asymptotically along the iteration axis. The theoretical findings are illustrated through simulation examples.