Neuro-fuzzy computing, which provides efficient information processing capability by devising methodologies and algorithms for modeling uncertainty and imprecise information, forms at this juncture, a key component of soft computing. An integrated neuro-fuzzy system is simply a fuzzy inference system trained by a neural networklearning algorithm. The learning mechanism fine-tunes the underlying...

Although neuro-fuzzy models and other similar modeis have great flexibility they also have drawbacks, especially for systems which have high uncertainty associated . This paper points out which are the major drawbacks of neuro-fuzzy models and proposes a methodology to design fault detectors in uncertain systems using neuro-fuzzy models. The coolíng coil of an air-conditioning system is used as...