Process diagnosis based on qualitative trend similarities using a sequence matching algorithm

Abstract

This paper focuses on process diagnosis based on symptoms described by qualitative trends extracted from signals. Main contributions are a new similarity algorithm between qualitative sequences and a generalised approach for on-line diagnosis based on that similarity measure. Process situations are identified by converting sensor time series into qualitative sequences and comparing them with those corresponding to known faulty states. Performance of this approach has been evaluated on line in a steam generator plant

This work has been supported by the research project DPI2009-07891 funded by the Spanish Government