Reprint of: Big data approach to batch process monitoring: Simultaneous fault detection and diagnosis using nonlinear support vector machine-based feature selection. (4th August 2018)

• Record Type: Journal Article
• Authors: Onel, Melis; Kieslich, Chris A.; Guzman, Yannis A.; Floudas, Christodoulos A.; Pistikopoulos, Efstratios N.
• Is Part Of: Computers & chemical engineering. Volume 116(2018)
• Page Start: 503
• Page End: 520
• Publication Date: 2018-08-04
• Subjects: Process monitoring -- Data-driven modeling -- Big data -- Feature selection -- Support vector machines; Chemical engineering -- Data processing -- Periodicals; 660.0285
• DOI: 10.1016/j.compchemeng.2018.10.016
• Languages: English
• ISSNs: 0098-1354

Abstract

Highlights: A novel data-driven framework using nonlinear Support Vector Machine-based feature selection is proposed for fault detection and diagnosis in batch processes. The proposed framework is applied on a comprehensive benchmark dataset comprising of 22, 600 batches with 15 faults, and normal operation. Fault and time-specific models are trained for simultaneous fault detection and diagnosis with three distinct time horizon approaches: one-step rolling, two-step rolling and evolving. Abstract: This paper presents a novel data-driven framework for process monitoring in batch processes, a critical task in industry to attain a safe operability and minimize loss of productivity and profit. We exploit high dimensional process data with nonlinear Support Vector Machine-based feature selection algorithm, where we aim to retrieve the most informative process measurements for accurate and simultaneous fault detection and diagnosis. The proposed framework is applied to an extensive benchmark data set which includes process data describing 22, 200 batches with 15 faults. We train fault and time-specific models on the pre-aligned batch data trajectories via three distinct time horizon approaches: one-step rolling, two-step rolling, and evolving which varies the amount of data incorporation during modeling. The results show that two-step rolling and evolving time horizon approaches perform superior to the other. Regardless of the approach, proposed framework provides a promising decision support tool for online simultaneous fault detection and diagnosis for batch processes.