Weighted part mutual information related component analysis for quality-related process monitoring. (April 2020)

• Record Type: Journal Article
• Authors: Wang, Yanwen; Zhou, Donghua; Chen, Maoyin; Wang, Min
• Is Part Of: Journal of process control. Volume 88(2020)
• Page Start: 111
• Page End: 123
• Publication Date: 2020-04
• Subjects: Process monitoring -- Quality-related fault detection -- Part mutual information -- Related component analysis -- Bayesian inference weighted fusion; Process control -- Periodicals; Fabrication -- Contrôle -- Périodiques; Process control; Periodicals; Electronic journals; 660.281
• DOI: 10.1016/j.jprocont.2020.03.001
• Languages: English
• ISSNs: 0959-1524

Abstract

Highlights: In this paper, we propose a weighted PMI based related component analysis (WPMI-RCA) method to search for the low-dimensional latent subspace of process variables that retains the maximal statistical associations with multiple quality variables. Considering different correlations with different quality variables, a new weighted fusion index is constructed by combining PMI with the weighted fusion of Bayesian inference. Process variables with strong relations to quality can be selected to remove the redundancy and filter the noise on the original measurement space. By transforming the axes in the measurement space, related components (RCs) with maximum correlation to quality are obtained by related component analysis (RCA) with the supervision of quality. Meaningful information hidden between process and quality data can be effectively captured. And the statistics in two orthogonal subspaces are defined for the quality-related process monitoring. The simulation results of the numerical example and the benchmark TEP show the superior monitoring performance of the presented method. The detection rates for quality-related faults can be improved, the nuisance detections of quality-unrelated anomalies can be reduced at the same time. Abstract: Industrial products have become the core of today's highly competitive international society, but quality-related faults happened in practical industrial processes heavily affect product quality. In this paper, we will consider the problem of the detection of quality-related faults. Inspired by part mutual information (PMI), we develop a process monitoring method called weighted PMI based related component analysis (WPMI-RCA). Firstly, combining PMI and Bayesian weighted fusion, process variables strongly related to quality are selected with the supervision of multi-quality indicators. Then, the selected variables are modeled by related component analysis (RCA) and thus orthogonal related components (RCs) containing the main information of quality variations can be obtained. The process data space can be divided into two subspaces and the monitoring statistics are developed for the quality-related fault detection. Finally, the validity of WPMI-RCA is demonstrated by a numerical example and the benchmark Tennessee Eastman process (TEP). The proposed method can improve the detection rates of quality-related faults and significantly reduce the nuisance detections. It may be helpful to improve the management efficiency for practical industrial processes.