Concurrent auto-regressive latent variable model for dynamic anomaly detection. (December 2021)

• Record Type: Journal Article
• Authors: Xu, Bo; Zhu, Qinqin
• Is Part Of: Journal of process control. Volume 108(2021)
• Page Start: 1
• Page End: 11
• Publication Date: 2021-12
• Subjects: Dynamic latent variable analysis -- Auto-regressive exogenous model -- Dynamic feature extraction -- Closed-loop control -- Quality-relevant monitoring; Process control -- Periodicals; Fabrication -- Contrôle -- Périodiques; Process control; Periodicals; Electronic journals; 660.281
• DOI: 10.1016/j.jprocont.2021.10.012
• Languages: English
• ISSNs: 0959-1524

Abstract

Abstract: With Industry 4.0, temporal relations between process and quality variables become increasingly complex. Some dynamic supervised learning algorithms are designed to extract their dynamic cross-relations, but auto-correlations among quality variables are rarely considered, which, however, can provide additional valuable information. This article proposes a novel dynamic auto-regressive latent variable model (DALVM) to capture both auto and cross correlations from high-dimensional time series data. DALVM is designed to maximize the covariance between current quality score and the weighted sum of past quality and process scores, and an auto-regressive exogenous inner model is developed for consistency purpose. Further, a concurrent anomaly detection system is developed based on DALVM, referred to as ConDALVM, which conducts subsequent decompositions in the extracted latent spaces. ConDALVM realizes a comprehensive monitoring for both static and dynamic anomalies in process and quality spaces. The superiority of the methods is demonstrated through a numerical simulation and two industrial processes. Highlights: Propose DALVM to model high dimensional time series data under closed loop control. Design a concurrent anomaly detection scheme with further decompositions on DALVM. Realize comprehensive static and dynamic monitoring in all extracted subspaces. Validate the efficacy of the proposed algorithms with three case studies.