Signed directed graph based modeling and its validation from process knowledge and process data

Abstract

This paper is concerned with the fusion of information from process data and process connectivity and its subsequent use in fault diagnosis and process hazard assessment. The Signed Directed Graph (SDG), as a graphical model for capturing process topology and connectivity to show the causal relationships between process variables by material and information paths, has been widely used in root cause and hazard propagation analysis. An SDG is usually built based on process knowledge as described by piping and instrumentation diagrams. This is a complex and experience-dependent task, and therefore the resulting SDG should be validated by process data before being used for analysis. This paper introduces two validation methods. One is based on cross-correlation analysis of process data with assumed time delays, while the other is based on transfer entropy, where the correlation coefficient between two variables or the information transfer from one variable to another can be computed to validate the corresponding paths in SDGs. In addition to this, the relationship captured by data-based methods should also be validated by process knowledge to confirm its causality. This knowledge can be realized by checking the reachability or the influence of one variable on another based on the corresponding SDG which is the basis of causality. A case study of an industrial process is presented to illustrate the application of the proposed methods.

Integration of process design and control: A review

“…Furthermore, signed directed graphs are often built bases on process connectivity described by piping and instrumentation diagrams and need validation. Yang et al, (2012) proposed two methods for validation of the extracted signed directed graphs, based on transfer entropy and cross correlation. Similarly, Maurya, et al, (2007) and Dong et al, (2010) combined qualitative trend analysis with signed directed graphs in order to improve the diagnostic resolution of their methods.…”

Progress in Root Cause and Fault Propagation Analysis of Large-Scale Industrial Processes

“…In (2), the PDF can be estimated by the kernel method [28,39], to fit any shape of the distributions.…”

“…model, while the SDG is only concerned with the qualitative trends. Since the exact model is often difficult to obtain for industrial processes, the SDG model is more widely used for its simplicity because it can be validated by process data [28].…”

A Causality Capturing Method for Diagnosis Based on Transfer Entropy by Analyzing Trends of Time Series

“…Existing techniques for causality capture can be generally divided into two types, namely, process knowledge based methods and process data based methods. The former includes adjacency and reachability matrices (Jiang et al, 2009), signed directed graph based methods (Vedam et al, 1997;Venkatasubramanian et al, 2000;Yang et al, 2012;Yang et al, 2013), and plant topology extraction by web language (Yim et al, 2006;Thambirajah et al, 2009), while the latter contains cross-correlation analysis, Granger causality analysis (Granger, 1969), Bayesian network learning (Cowell et al, 1999), transfer entropy (Schreiber, 2000) (TE), and other entropy based methods (Doyle et al, 1993a, b). Besides, to achieve better diagnosis performance, especially when abnormal situations are associated with unknown faults or multiple faults, integrated methods that combine process data analysis with process knowledge extraction are proposed and demonstrated to be quite effective (Chiang et al, 2003).…”