Premium
A method of fault diagnosis in adaptive control systems by using fuzzy inference
Author(s) -
Kumararu Kousuke
Publication year - 1992
Publication title -
electrical engineering in japan
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.136
H-Index - 28
eISSN - 1520-6416
pISSN - 0424-7760
DOI - 10.1002/eej.4391120808
Subject(s) - fault (geology) , adaptive neuro fuzzy inference system , control theory (sociology) , inference , fuzzy logic , fuzzy inference system , fuzzy control system , computer science , scheme (mathematics) , data mining , control engineering , engineering , artificial intelligence , control (management) , mathematics , mathematical analysis , seismology , geology
An on‐line fault diagnosis scheme for adaptive controlled systems designed by a self‐tuning approach is proposed. A physical parameter change in the controlled system can effectively be detected by using Kullback Discrimination Information (KDI) as an index for model discrimination. In the adaptive controlled system, parameter changes may occur under the normal operation, as well as under a failed situation. In order to decide whether the detected system parameter change really means a fault occurrence or not, a fuzzy inference approach to fault diagnosis is considered. Some appropriate membership functions which describe fuzzy events of the fault are constructed to implement the fuzzy inference. In this way, useful knowledge about fault modes obtained from, e.g., experts, can be introduced into the model‐based diagnosis technique. Simulation studies on a second order damped oscillator have been carried out to demonstrate the effectiveness of the method.