Anomaly detection in time series data using a combination of wavelets, neural networks and Hilbert transform

Stratis Kanarachos; Jino Mathew; Alexander Chroneos; Michael E. Fitzpatrick

doi:10.1109/IISA.2015.7388055

Anomaly detection in time series data using a combination of wavelets, neural networks and Hilbert transform

Stratis Kanarachos, Jino Mathew, Alexander Chroneos, Michael E. Fitzpatrick

College of Engineering, Environment & Science

Research output: Contribution to conference › Paper

21 Citations (Scopus)

531 Downloads (Pure)

Abstract

Real time detection of anomalies is crucial in structural health monitoring applications as it is used for early detection of structural damage and to identify abnormal operating conditions that can shorten the life of operating structures. A new signal processing algorithm for detecting anomalies in time series data is proposed in this study. The algorithm is expressed as a combination of wavelet analysis, neural networks and Hilbert transform in a sequential manner. The algorithm has been evaluated for a number of benchmark tests, commonly used in the literature, and has been found to perform robustly.

Original language	English
DOIs	https://doi.org/10.1109/IISA.2015.7388055
Publication status	Published - 2015
Event	6th IEEE International Conference on Information, Intelligence, Systems and Applications - Corfu, Greece Duration: 6 Jul 2015 → 8 Jul 2015

Conference

Conference	6th IEEE International Conference on Information, Intelligence, Systems and Applications
Abbreviated title	IISA2015
Country/Territory	Greece
City	Corfu
Period	6/07/15 → 8/07/15

Bibliographical note

© 2015 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.

Keywords

anomaly detection
wavelets
neural networks
Hilbert

Access to Document

10.1109/IISA.2015.7388055

Anomaly detection

Cite this

@conference{8bd8e182d13f4fedb03e3dc549ad1852,

title = "Anomaly detection in time series data using a combination of wavelets, neural networks and Hilbert transform",

abstract = "Real time detection of anomalies is crucial in structural health monitoring applications as it is used for early detection of structural damage and to identify abnormal operating conditions that can shorten the life of operating structures. A new signal processing algorithm for detecting anomalies in time series data is proposed in this study. The algorithm is expressed as a combination of wavelet analysis, neural networks and Hilbert transform in a sequential manner. The algorithm has been evaluated for a number of benchmark tests, commonly used in the literature, and has been found to perform robustly. ",

keywords = "anomaly detection, wavelets, neural networks, Hilbert",

author = "Stratis Kanarachos and Jino Mathew and Alexander Chroneos and Fitzpatrick, {Michael E.}",

note = "{\textcopyright} 2015 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.; 6th IEEE International Conference on Information, Intelligence, Systems and Applications, IISA2015 ; Conference date: 06-07-2015 Through 08-07-2015",

year = "2015",

doi = "10.1109/IISA.2015.7388055",

language = "English",

}

TY - CONF

T1 - Anomaly detection in time series data using a combination of wavelets, neural networks and Hilbert transform

AU - Kanarachos, Stratis

AU - Mathew, Jino

AU - Chroneos, Alexander

AU - Fitzpatrick, Michael E.

N1 - © 2015 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.

PY - 2015

Y1 - 2015

N2 - Real time detection of anomalies is crucial in structural health monitoring applications as it is used for early detection of structural damage and to identify abnormal operating conditions that can shorten the life of operating structures. A new signal processing algorithm for detecting anomalies in time series data is proposed in this study. The algorithm is expressed as a combination of wavelet analysis, neural networks and Hilbert transform in a sequential manner. The algorithm has been evaluated for a number of benchmark tests, commonly used in the literature, and has been found to perform robustly.

AB - Real time detection of anomalies is crucial in structural health monitoring applications as it is used for early detection of structural damage and to identify abnormal operating conditions that can shorten the life of operating structures. A new signal processing algorithm for detecting anomalies in time series data is proposed in this study. The algorithm is expressed as a combination of wavelet analysis, neural networks and Hilbert transform in a sequential manner. The algorithm has been evaluated for a number of benchmark tests, commonly used in the literature, and has been found to perform robustly.

KW - anomaly detection

KW - wavelets

KW - neural networks

KW - Hilbert

U2 - 10.1109/IISA.2015.7388055

DO - 10.1109/IISA.2015.7388055

M3 - Paper

T2 - 6th IEEE International Conference on Information, Intelligence, Systems and Applications

Y2 - 6 July 2015 through 8 July 2015

ER -

Anomaly detection in time series data using a combination of wavelets, neural networks and Hilbert transform

Abstract

Conference

Bibliographical note

Keywords

Access to Document

Fingerprint

Cite this