Diagnosis of power plant faults using qualitative models and heuristic rules

ABSTRACT

This paper presents results obtained in an AI research effort in the industrial field of Nuclear Power Plants (NPP): malfunction diagnosis of the Emergency Feedwater System (EFWS) of a NPP. An expert system was developed which utilizes qualitative techniques for modeling the system and heuristic rules for generating causal explanations of an observed malfunction. The operation of the system, the model and the global inference mechanism are discussed. Another purpose of the paper is to present the pro and cons of this approach in this application of industrial relevance.

REFERENCES

• 1: Bobrow, D.G. (Ed.). "Special Volume on Qualitative Reasoning about Physical Systems". Artificial Intelligence 2d (1-3), 1984.
• 2: Johan de Kleer, An assumption-based TMS, Artificial Intelligence, v.28 n.2, p.127-162, March 1986 [doi>10.1016/0004-3702(86)90080-9]
• 3: J de Kleer , B C Williams, Diagnosing multiple faults, Artificial Intelligence, v.32 n.1, p.97-130, April 1987 [doi>10.1016/0004-3702(87)90063-4]
• 4: Herbert, M.R., and G.H. Williams. "An Examination of Qualitative Plant Modelling as a Basis for Knowledge-Based Operator Aids in Nuclear Power Stations". Proceedings of the Unicom Seminar on Expert Systems and Optimization in Process Control, London, December, 1985.
• 5: Herbert, M.R., and G.H. Williams. "An initial evaluation of the detection and diagnosis of power plant faults using deep knowledge representation of physical behaviour". Expert Systems, vol. 4, no. 2, May 1987.
• 6: Kasahara, T., Kato, K., and T. Ito. "Maintenance Work Scheduling Aid for Nuclear Power Plants". International Workshop on Artificial Intelligence for Industrial Applications, pp. 161-166, 1988.
• 7: Lewis, E.E. "Loss-of-coolant accidents". In J. Wiley (Ed.), Nuclear Power Reactor Safety, Vol. I, 1st ed. John Wiley & Sons, Inc., New York. pp. 407- 423, 1977.
• 8: Obreja, I., and G. Friedrich. "Model-based decision tree generation for diagnosis and measurement selection". Proceedings of the IFAC/IFIP SAFE- COMP 1989 Workshop. Vienna, 1989 Dec 05-07, pp. 109~115.
• 9: Parsay, K., and K.Y. Liu. "An Expert System Structure for Automatic Fault Tree Generation for Emergency Feedwater Systems for Nuclear Power Plants", IEEE, pp. 25-30, 1987.
• 10: Quinlan, J.R. "Learning efficient classification procedures and their application to chess end games", in R.S. Michalski, J.C. Carbonell and T.M. Mitchell (Eds.), Machine Learning- An Artificial Intelligence Approach, chapter 15, pp. 463- 482, Morgan Kaufmann Publishers, Inc., 1983.
• 11: R Reiter, A theory of diagnosis from first principles, Artificial Intelligence, v.32 n.1, p.57-95, April 1987 [doi>10.1016/0004-3702(87)90062-2]
• 12: F. Van Harmelen , A. Bundy, Explanation-based generalisation = partial evaluation, Artificial Intelligence, v.36 n.3, p.401-412, October 1988 [doi>10.1016/0004-3702(88)90088-4]
• 13: "Review of the Seabrook Units 1 and 2 Auxiliary Feedwater System Reliability Analysis". NUREG- 3531, 1984.

INDEX TERMS

Primary Classification:
  J. Computer Applications
  J.2 PHYSICAL SCIENCES AND ENGINEERING
      Subjects: Electronics

Additional Classification:
  I. Computing Methodologies
  I.2 ARTIFICIAL INTELLIGENCE
      I.2.1 Applications and Expert Systems
          Subjects: Industrial automation
      I.2.6 Learning
          Subjects: Knowledge acquisition
      I.2.8 Problem Solving, Control Methods, and Search
          Subjects: Heuristic methods

General Terms:
Algorithms, Design, Performance, Reliability