2008 年 2008 巻 7-3 号 p. 691-696
The goal of this paper is to detect internal leakages created by seal faults and also identify which seal is damaged from water hydraulic proportional valve using vibration and pressure signals which are analyzed using different data analysis methods. In this study water hydraulic spool valve is studied in a test system during extending and retracting strokes of cylinder while vibration and pressure signals from the valve are measured in normal and different fault situations. Feature extraction is performed using descriptive statistics and wavelet analysis to identify the most influential variables from the measured signals which are then used to classify the state of the system using Self-Organizing Maps (SOM).