A new auto-encoder-based dynamic threshold to reduce false alarm rate for anomaly detection of steam turbines. (1st March 2022)

• Record Type: Journal Article
• Authors: Ko, Jin Uk; Na, Kyumin; Oh, Joon-Seok; Kim, Jaedong; Youn, Byeng D.
• Is Part Of: Expert systems with applications. Volume 189(2022)
• Publication Date: 2022-03-01
• Subjects: Anomaly detection -- Deep learning -- Ensemble denoising auto-encoder -- Dynamic threshold -- Steam turbine; Expert systems (Computer science) -- Periodicals; Systèmes experts (Informatique) -- Périodiques; Electronic journals; 006.33
• DOI: 10.1016/j.eswa.2021.116094
• Languages: English
• ISSNs: 0957-4174

Abstract

Highlights: The proposed method can model the normal data with denoising and ensemble technique. Dynamic threshold is newly developed to minimize false alarms in anomaly detection. Sensitivity is newly defined to identify condition parameters related to an anomaly. New metrics are defined to validate the anomaly detection performance. Abstract: This study proposes an ensemble denoising auto-encoder-based dynamic threshold (EDAE-DT) to overcome the false alarm issue in anomaly detection. The proposed ensemble denoising auto-encoder can model the normal condition well through a denoising task and an ensemble technique. The dynamic threshold sets a time-varying threshold that considers the variation of normal data. Performance metrics for anomaly detection are newly proposed to quantitatively verify the performance. A new sensitivity is defined from the dynamic threshold to identify which signal is related to the change that arises due to an anomaly. The diagnostic performance of the proposed approach is compared using metrics for classification and a confusion matrix. Validation results, which examined thermal power plant datasets, show that the proposed modeling method outperforms both the auto-encoder and denoising auto-encoder approaches. Additionally, the proposed method can significantly reduce the false alarm rate, as compared to conventional methods, while detecting anomalies faster than experts. The anomaly-related signals are identified successfully through the newly defined sensitivity. Finally, the diagnositc results demonstrate that the proposed approach is more accurate than conventional methods.