Application of fuzzy-neural networks in fault diagnosis
Yimin Li, HU Shou-song · Systems engineering and electronics · 2005
Fault diagnosis become more and more important to keep the whole plant safe and reliable. The model-based fault detection and diagnosis approach can not satisfy the increasing complexity and nonlinear. The causalities between the faults and the symptoms are quite difficult to be modeled analytically, so a free model approach is more suitable to acquire represent and process the diagnostic knowledge base. The combinatorial forms of several fuzzy-neural networks with different structures are analyzed. The direct fuzzy-neuron network (FNN), B-spline function network、 RBFNN and some extending fuzzy-neural networks applications in FDD are studied. And a general fault diagnosis model and method is proposed.