Generation of Signed Directed Graphs Using Functional Models

Abstract Intelligent fault diagnosis systems can be a major aid to human operators charged with the high-level control of industrial plants. Such systems aim for high diagnostic accuracy while retaining the ability to produce results that can be interpreted by human experts on site. Signed directed graphs have been shown to be a viable method for plant-wide diagnosis that can incorporate both quantitative information about the process condition as well as qualitative information about the system topology and the functions of its components. Their range of application in industrial settings has been limited due to difficulties regarding the interpretation of results and consistent graph generation. This contribution addresses these issues by proposing an automated generation of signed directed graphs of industrial processes in the chemical, petroleum and nuclear industries using Multilevel Flow Modeling; a functional modeling method designed for operator support. The approach is demonstrated through a case study conducted on the Tennessee Eastman Process, showing that Multilevel Flow Modeling can facilitate a consistent modeling process for signed directed graphs. Finally, the resulting benefits regarding qualitative reasoning for plant-wide diagnosis are discussed.

[1]  Petr Ekel,et al.  A new fault classification approach applied to Tennessee Eastman benchmark process , 2016, Appl. Soft Comput..

[2]  Sirish L. Shah,et al.  Signed directed graph based modeling and its validation from process knowledge and process data , 2012, Int. J. Appl. Math. Comput. Sci..

[3]  JayHyung Lee,et al.  Nonlinear modeling and state estimation for the Tennessee Eastman challenge process , 1995 .

[4]  Ole Ravn,et al.  Consequence Reasoning in Multilevel Flow Modelling , 2013, IFAC HMS.

[5]  Elwood S. Buffa,et al.  Graph Theory with Applications , 1977 .

[6]  Tetsuo Tomiyama,et al.  A review of function modeling: Approaches and applications , 2008, Artificial Intelligence for Engineering Design, Analysis and Manufacturing.

[7]  Lind Morten,et al.  An introduction to multilevel flow modeling , 2011 .

[8]  Raghunathan Rengaswamy,et al.  A review of process fault detection and diagnosis: Part II: Qualitative models and search strategies , 2003, Comput. Chem. Eng..

[9]  Xinxin Zhang.  Assessing Operational Situations. , 2015 .

[10]  Ping Zhang,et al.  A comparison study of basic data-driven fault diagnosis and process monitoring methods on the benchmark Tennessee Eastman process , 2012 .

[11]  E. F. Vogel,et al.  A plant-wide industrial process control problem , 1993 .

[12]  Mark A. Kramer,et al.  A rule‐based approach to fault diagnosis using the signed directed graph , 1987 .

[13]  Yu Wang,et al.  An intelligent fault diagnosis system for process plant using a functional HAZOP and DBN integrated methodology , 2015, Eng. Appl. Artif. Intell..

[14]  Poong Hyun Seong,et al.  Enhanced reasoning with multilevel flow modeling based on time-to-detect and time-to-effect concepts , 2018 .

[15]  Fredrik Dahlstrand,et al.  Improving expressional power and validation for multilevel flow models , 2004, J. Intell. Fuzzy Syst..

[16]  Dazi Li,et al.  A hybrid fault diagnosis method based on fuzzy signed directed graph and neighborhood rough set , 2017, 2017 6th Data Driven Control and Learning Systems (DDCLS).

[17]  Raghunathan Rengaswamy,et al.  A review of process fault detection and diagnosis: Part I: Quantitative model-based methods , 2003, Comput. Chem. Eng..

[18]  Douglas H Rothenberg,et al.  Alarm Management for Process Control: A Best-Practice Guide for Design, Implementation, and Use of Industrial Alarm Systems , 2009 .

[19]  Daniel A. Crowl,et al.  Chemical Process Safety: Fundamentals with Applications , 2001 .

[20]  Qunxiong Zhu,et al.  A Multilogic Probabilistic Signed Directed Graph Fault Diagnosis Approach Based on Bayesian Inference , 2014 .

[21]  Fabrice Druaux,et al.  Indirect neural control for plant-wide systems: Application to the Tennessee Eastman Challenge Process , 2010, Comput. Chem. Eng..

[22]  Hao Ye,et al.  Statistical root cause analysis of novel faults based on digraph models , 2013 .

[23]  Raghunathan Rengaswamy,et al.  A Signed Directed Graph and Qualitative Trend Analysis-Based Framework for Incipient Fault Diagnosis , 2007 .

[24]  Raghunathan Rengaswamy,et al.  A review of process fault detection and diagnosis: Part III: Process history based methods , 2003, Comput. Chem. Eng..