Fuzzy logic: a modelling tool for transient diagnostics

The identification and classification of signal trends is a fundamental phase for the reliable monitoring and timely diagnosing of nuclear components. The main target is the early detection of the onset of a transient and the classification of its causes. Given the safety and economical importance of the problem, various approaches have been investigated and applied for trend identification, and many efforts are still devoted to the improvement of the results so far obtained. In this paper, a new fuzzy-logic based method for classification of transients' initiating events is proposed. The if-then rules, which constitute the heart of the model, are inferred from the available input-output signal data. The method is applied to the early classification of the causes of transients in a steam generator of a Pressurized Water Reactor (PWR). Based on the measured signals, the forcing function responsible for the transient is readily classified.