Fault Diagnosis System Development for Rotating Electrical Machinery

• Main Authors: Jen-Chen Tsai, 蔡鎮丞
• Other Authors: 黃坤元
• Format: Others
• Language: zh-TW
• Published: 2011
• Online Access: http://ndltd.ncl.edu.tw/handle/00907609306541719105

碩士 === 正修科技大學 === 電機工程研究所 === 99 === The thesis proposes a artificail neural network(ANN) for rotor machinery fault diagosis system. The fault detection is important for the the electrical rotor machinery. The breakdown of the rotor electrical machinery, especially the important machines, will bring about interruption of production and reduce profits. If there is a predictive fault diagnosis system, the situation of the outage can be avoided. However, the components of the rotor machinery are more complex and more sophisticated recently. These causes resulting in the machine vibration problems become more various and complicated. The mechanical vibration signal is a major parameter for the predictive maintenance system. The subject draws much attention in the predictive detection research. However, the electrical signal is also an important response when the electrical rotor machinery is breakdown. The thesis propose a excellent fault detection method, which is not only vibration signal considered but also electrical signal. The ANN is used to forecast for the fault diagnosis system in the study. The operation situation of machine can be detected by the parameters of operational pressure, temperature, and vibration. The thesis, also proposes the input current of machinery for more precisely diagnosing the equipment operation situation. These parameters are used to set the training patterns for the ANN to develop the fault diagnosis systems. Keywords: rotor electrical machinery, fault diagnosis, artificial neural network