Fault Detection and Isolation using Correspondence Analysis

Authors:	Gudi Ravindra, IIT Bombay, India Detroja Ketan, IIT Bombay, India Patwardhan Sachin, IIT Bombay, India
Topic:	6.4 Safeprocess
Session:	Signal Based Fault Detection and Isolation
Keywords:	Correspondence Analysis, Generalized Singular Value Decomposition, Fault Detection and Isolation

Abstract

In this paper, a new approach to Fault Detection and Diagnosisthat is based on Correspondence Analysis is proposed. Correspondence analysisis a powerful multivariate technique based on the generalized singular valuedecomposition. The advantage of using correspondence analysis is that it depictsrows as well as columns as points in the dual lower dimensional vector space.Correspondence analysis has been shown to capture association between variousfeatures and events quite e¤ectively. In this paper, the correspondence analysisapproach is used for Fault Detection and Diagnosis (FDD) and is validated onrepresentative process systems.