Comparison of structural analysis and principle component analysis for leakage prediction on superheater in boiler

<span>Leakage is one of the failures which commonly happens in boiler operation. Moreover, a continuous unsettled anomaly could lead to leakage failure. An algorithm has been developed predict failure, consisting three general procedures: feature selection, followed by hierarchical clustering, and naïve Bayes classification. The clustering changes unlabeled data into labeled data, classification calculates probability justify occurrence. Meanwhile, this research focused on effect selection method result prediction. Two different methods, namely structural analysis principal component (PCA), were deployed separately then compared. showed that prediction using gave 13 hours 40 minutes time, PCA 25 time. However, caused more false alarms than with analysis, only triggered five week before leakage. offered better traceability understand occurrence.</span>