Robust Fault Diagnosis for the Chemical Plant Based on the Qualitative Model

It is a weak point of the algorithm of fault diagnosis utilizing the signed digraph with delay that we cannot quantitatively estimate the affect of errors in delays on its diagnostic accuracy. In this paper, an algorithm of fault diagnosis is proposed in order to give the values of membership function to the candidates for the origin of failure by representing delays in fuzzy numbers. And a concept of the information content of the diagnostic result is presented to make a quantitative estimation of the diagnostic accuracy. Furthermore, experiments with the use of a mini-plant demonstrate that the amount of information of the diagnostic result by the proposed algorithm is little affected by the errors in delays.