A multivariate based event detection method and performance comparison with two baseline methods. (1st September 2015)

• Record Type: Journal Article
• Authors: Liu, Shuming; Smith, Kate; Che, Han
• Is Part Of: Water research. Volume 80(2015)
• Page Start: 109
• Page End: 118
• Publication Date: 2015-09-01
• Subjects: Contaminant classification -- Conventional sensor -- Early warning system -- Euclidean distance -- Pearson correlation -- Water quality; ANN artificial neural network -- ARMA autoregressive moving average -- CIE contaminant injection experiment -- EWS early warning system -- FAR false alarm rate -- FN false negative -- FP false positive -- LPF linear prediction filters -- MED multivariate Euclidean distance -- ORP oxidation reduction potential -- PE Pearson correlation Euclidean distance-based method -- PD probability of detection -- READiw real-time event adaptive detection, identification and warning -- ROC receiver operating characteristic -- SCADA supervisory control and data acquisition -- SVM support vector machine -- TN true negative -- TOC total organic carbon -- TP true positive; Water -- Pollution -- Research -- Periodicals; 363.7394
• DOI: 10.1016/j.watres.2015.05.013
• Languages: English
• ISSNs: 0043-1354

Abstract

Abstract: Early warning systems have been widely deployed to protect water systems from accidental and intentional contamination events. Conventional detection algorithms are often criticized for having high false positive rates and low true positive rates. This mainly stems from the inability of these methods to determine whether variation in sensor measurements is caused by equipment noise or the presence of contamination. This paper presents a new detection method that identifies the existence of contamination by comparing Euclidean distances of correlation indicators, which are derived from the correlation coefficients of multiple water quality sensors. The performance of the proposed method was evaluated using data from a contaminant injection experiment and compared with two baseline detection methods. The results show that the proposed method can differentiate between fluctuations caused by equipment noise and those due to the presence of contamination. It yielded higher possibility of detection and a lower false alarm rate than the two baseline methods. With optimized parameter values, the proposed method can correctly detect 95% of all contamination events with a 2% false alarm rate. Graphical abstract: Highlights: New contamination detection method developed for use in early warning systems. Compares Euclidean distances of correlation indicators from water quality sensors. The method differentiates well between equipment noise and real contamination. It can detect 95% of contamination events correctly with a 2% false alarm rate. It significantly outperforms two conventional detection methods.