This paper is to present a novel design of the Iterative Learning Observer (ILO) for the purpose of fault estimation. With the new design of ILO, initial oscillation of the fault estimate is dramatically reduced. The learning accuracy is accordingly improved. Convergence of the ILO to the system is proved and simulation study is conducted. It shows that accurate fault estimation can be performed by the new ILO.

This content is only available via PDF.
You do not currently have access to this content.