A generalized change detection problem

Author

Nikiforov, Igor V.

Author_Institution

Inst. of Control Sci., Moscow, Russia

Volume

41

Issue

1

fYear

1995

fDate

1/1/1995 12:00:00 AM

Firstpage

171

Lastpage

187

Abstract

The purpose of this paper is to give a new statistical approach to the change diagnosis (detection/isolation) problem. The change detection problem has received extensive research attention; however, the change isolation problem has, for the most part, been ignored. We consider a stochastic dynamical system with abrupt changes and investigate the multiple hypotheses extension of Lorden´s (1971) results. We introduce a joint criterion of optimality for the detection/isolation problem and then design a change detection/isolation algorithm. We also investigate the statistical properties of this algorithm. We prove a lower bound for the criterion in a class of sequential change detection/isolation algorithms. It is shown that the proposed algorithm is asymptotically optimal in this class. The theoretical results are applied to the case of additive changes in linear stochastic models

Keywords

optimisation; signal detection; statistical analysis; stochastic processes; Lorden´s results; abrupt changes; additive changes; asymptotically optimal algorithm; change diagnosis; change isolation problem; generalized change detection problem; joint optimality criterion; linear stochastic models; lower bound; multiple hypotheses extension; sequential change detection algorithms; sequential change isolation algorithms; statistical approach; stochastic dynamical system; Change detection algorithms; Fault detection; Fault diagnosis; Image edge detection; Radar detection; Signal processing; Signal processing algorithms; Sonar detection; Stochastic processes; Stochastic systems;

fLanguage

English

Journal_Title

Information Theory, IEEE Transactions on

Publisher

ieee

ISSN

0018-9448

Type

jour

DOI

10.1109/18.370109

Filename

370109