Quantile index for gradual and abrupt change detection from CFB boiler sensor data in online settings

A. Maslov, M. Pechenizkiy, T. Kärkkäinen, M. Tähtinen

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

2 Citations (Scopus)
3 Downloads (Pure)

Abstract

In this paper we consider the problem of online detection of gradual and abrupt changes in sensor data having high levels of noise and outliers. We propose a simple heuristic method based on the Quantile Index (QI) and study how robust this method is for detecting both gradual and abrupt changes with such data. We evaluate the performance of our method on the arti¿cially generated and real datasets that represent di¿erent operational settings of a pilot circulating ¿uidized bed (CFB) reactor and CFB cold model. Our experiments suggest that QI can be used for designing very simple yet e¿ective methods for gradual change detection in the noisy sensor data. It can be also used for detecting abrupt changes in the data unless they occur too often one after another.
Original languageEnglish
Title of host publicationProceedings of the Sixth International Workshop on Knowledge Discovery from Sensor Data (SensorKDD'12, Beijing, China, August 12, 2012)
Place of PublicationNew York
PublisherAssociation for Computing Machinery, Inc
Pages25-33
ISBN (Print)978-1-4503-1554-8
DOIs
Publication statusPublished - 2012
Eventconference; Sixth International Workshop on Knowledge Discovery from Sensor Data; 2012-08-12; 2012-08-12 -
Duration: 12 Aug 201212 Aug 2012

Conference

Conferenceconference; Sixth International Workshop on Knowledge Discovery from Sensor Data; 2012-08-12; 2012-08-12
Period12/08/1212/08/12
OtherSixth International Workshop on Knowledge Discovery from Sensor Data

Fingerprint

Dive into the research topics of 'Quantile index for gradual and abrupt change detection from CFB boiler sensor data in online settings'. Together they form a unique fingerprint.

Cite this