A Comparison of a Neural Network and an Observer Approach for Detecting Faults in a Benchmark System

  • D. N. Shields
  • S. Du
  • E. Gaura
Conference paper


The detection of faults is considered for a class of nonlinear systems. A fault detection observer approach is compared to a neural network approach. Both approaches are applied to a an experimental ( benchmark) three-tank system.


Fault Detection Neural Network Approach Benchmark System Observer Approach Fault Diagnosis System 
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]
    Frank, P.M. (1994). On-line Fault Detection in Uncertain Nonlinear Systems Using Diagnostic Observers: A Survey. Int.J.Systems Sci, Vol. 25, pp. 2129–2154.CrossRefMATHGoogle Scholar
  2. [2]
    Garcia, F. J., V. Izquierdo, L. de Miguel, J. Peran (1997). Fuzzy Identification of Systems and its Applications to Fault Diagnosis Systems. IFAC Symposium on Fault Detection, Supervision and Safety for Technical Processes “SA FEPR 0-CESS’97”, Kingston Upon Hull, Vol.2, 705–712.Google Scholar
  3. [3]
    Han, Z., P. M. Frank(1997). Physical Parameter Estimation Based FDI with Neural Networks. IFAC Symposium on Fault Detection, Supervision and Safety for Technical Processes “SAFEPROCESS’97”, Kingston Upon Hull, Vol.1, 294–299.Google Scholar
  4. [4]
    Shields, D.N. and S. Daley (1998). A Quantitative Fault Detection Method for a Class of Nonlinear Systems. Trans. Inst. MC 20(3), 125–133.CrossRefGoogle Scholar
  5. [5]
    Yu, D. and D.N. Shields (1996). Bilinear Fault Detection Observer and its Application to a Hydraulic System. Int. Jnl. of Control, Vol. 64, pp. 1023–1047.MathSciNetCrossRefMATHGoogle Scholar

Copyright information

© Springer-Verlag Wien 2001

Authors and Affiliations

  • D. N. Shields
  • S. Du
  • E. Gaura
    • 1
  1. 1.Maths-MISCoventry UniversityCoventryUK

Personalised recommendations