Rotor Crack Detection by Using Multi-vibration Signal from The Basement

Rotor crack detection method by using multi-vibration signals gathered at the basements is presented in this paper. The finite element software ANSYS is applied for analyzing the vibration response characteristics of the basement to determine sensor configuration. Feature fusion for time-domain statistics of multiple sensors is performed by using the support vector machine to diagnose the depth of crack rotor fault. Test analysis indicates that this method is fast and accurate to detect rotor crack by fusing the multi-vibration signal of the basement, and also supply a new approach for rotor fault diagnosis.