Diagnosis of Unanticipated Plant Component Faults in a Portable Expert System

  • Jaques Reifman
  • Thomas Y. C. Wei
Part of the Studies in Fuzziness and Soft Computing book series (STUDFUZZ, volume 38)


We describe the first-principles-based PRODIAG expert system for on-line plant-level diagnosis of component faults in thermal-hydraulic processes. This diagnostic system combines the concepts of fundamental physical principles and function-oriented diagnosis in a qualitative reasoning framework and structures these concepts into three independent knowledge bases. PRODIAG has the unique ability to diagnose unanticipated (unforeseen) component faults and can be ported across different processes/plants through modifications of only input data files containing the appropriate process layout information. Simulation tests for two plant systems with transient data generated with the Braidwood Nuclear Power Plant full-scope training simulator confirm the unique capabilities of PRODIAG.


Heat Exchanger Open Valve Component Fault Faulty Component Mass Imbalance 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    W. R. Nelson, “REACTOR: An Expert System for Diagnosis and Treatment of Nuclear Reactor Accidents,” Proc. National Conference on Artificial Intelligence, AAAI, August 18–20, 1982, Pittsburgh, PA, pp. 296–301, (1982).Google Scholar
  2. 2.
    J. Reifman, “Survey of Artificial Intelligence Methods for Detection and Identification of Component Faults in Nuclear Power Plants,” Nucl. Technol., Vol. 119, pp. 76–97 (1997).Google Scholar
  3. 3.
    J. De Kleer and J. S. Brown, “A Qualitative Physics Based on Confluences,” AI, Vol. 24, pp. 7–83 (1984).Google Scholar
  4. 4.
    J. Reifman and T. Y. C. Wei, “PRODIAG - Dynamic Qualitative Analysis for Process Fault Diagnosis,” Proc. Ninth Power Plant Dynamics, Control and Testing Symposium, Knoxville, Tennessee, May 24–26, 1995, Vol. 1, pp. 40.01–40.15, B. R. Upadhyaya, E. M. Katz, and T. W. Kerlin, Eds., The University of Tennessee, Knoxville, Tennessee (1995).Google Scholar
  5. 5.
    J. Reifman and T. Y. C. Wei, “PRODIAG: A Process-Independent Transient Diagnostic System-I: Theoretical Concepts,” Nucl. Sci. Eng., Vol. 131, pp. 329–347 (1999).Google Scholar
  6. 6.
    F. E. FINCH and M. A. KRAMER, “Narrowing Diagnostic Focus Using Functional Decomposition,” AIChE J., Vol. 34, 25–36 (1988).CrossRefGoogle Scholar
  7. 7.
    J. A. HASSBERGER and J. C. LEE, “Macroscopic Mass and Energy Balance for Nuclear Plant Diagnostics Using Fuzzy Logic,” Proc. Topical Meeting on Artificial Intelligence and Other Innovative Computer Applications in the Nuclear Industry, Snowbird, Utah, August 31-September 2, 1987, pp. 539–546, M. C. Majumdar, D. Majumdar, and J. I. Sacket, Eds., Plenum Press, New York (1988).Google Scholar
  8. 8.
    I. Bratko, Prolog Programming for Artificial Intelligence, Addison-Wesly (1986).Google Scholar
  9. 9.
    Braidwood Chemical and Volume Control System - System Description, Commonwealth Edison Company, Braidwood, Illinois (1990).Google Scholar
  10. 10.
    J. Reifman and T. Y. C. Wei, “PRODIAG: A Process-Independent Transient Diagnostic System-II: Validation Tests,” Nucl. Sci. Eng., Vol. 131, pp. 348–369 (1999).Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2000

Authors and Affiliations

  • Jaques Reifman
    • 1
  • Thomas Y. C. Wei
    • 1
  1. 1.Argonne National LaboratoryArgonneUSA

Personalised recommendations