A hierarchical fuzzy neural network approach for multiple fault diagnosis

A hierarchical structure of fuzzy neural networks for multiple abrupt and incipient fault diagnosis, is presented. It is discussed how to train the fuzzy neural network with single fault symptoms, in order to cope with diagnosis of multiple faults. The fuzzy neural network compresses the online measurement data and, hence, training effort is eased. A continuous binary distillation column is used as test bed of the overall system