Skip to Main Content
For reducing the economic losses of hydraulic generator units, fault detection and diagnosis is fundamental. A technique is presented that uses pattern recognition and cluster analysis. Faults of hydraulic generator units are classified from the characteristic parameters using statistical analysis methods. A characteristic parameter matrix of standard faults has been developed for identifying the vibration faults. The cluster technique is carried out in a real case. Results demonstrated that the proposed method is a good candidate to be used as an online diagnosis tool for hydraulic generator units.