Time-varying norm optimal iterative learning identification

In this paper, we focus on improving the performance of an Iterative Learning Identification (ILI) algorithm for identifying discrete, Single-Input Single-Output (SISO), Linear Time- Varying (LTV) plants that are able to repeat their trajectories. The identification learning laws are determined through an optimization framework, which is similar in nature to the design of norm optimal Iterative Learning Control (ILC). The ILI algorithm has been previously demonstrated to be capable of tracking rapid parameter changes. However, when it is applied to systems with noise, it results in high frequency parameter fluctuation around their true values. This paper suggests a time-varying ILI technique to improve the steady state estimation while maintaining the ILI's ability to track rapid parameter changes.