期刊名称：Lecture Notes in Engineering and Computer Science
印刷版ISSN：2078-0958
电子版ISSN：2078-0966
出版年度：2018
卷号：2231&2232
页码：303-307
出版社：Newswood and International Association of Engineers
摘要：This paper discusses the diagnosis of shorted-turn
faults in the electrical machine using Neural Networks (NN).
This leads to a design process of a work-flow for the NN. The
work-flow has three stages: data acquisition, training algorithm
and diagnosis and detection of machine condition. Samples
data of electrical machine in healthy and shorted-turn fault
conditions were collected by interfacing data acquisition device
with a computer laboratory. A two-layer feed-forward network
with back-propagation algorithm is created and configured with
data collected for NN training. The network model gives a high
correlation coefficient of R = 0.9992, R = 0.99917 and R =
0.99923 in the training, validation and test phase respectively
as well as the overall correlation which is R = 0.9992. This
connotes that the NN model gives a high correlation coefficient
between predicted outputs (NN) and targets (Fault Index (FI)).
Using the NN model, the healthy and shorted-turn electrical
machine are predicted correctly and this is compared with the
diagnosis done using FI. Thus, with an NN, a robust and reliable
method to diagnose shorted-turn fault in the electrical machine
can be achieved.
关键词：electrical machine; fault diagnosis; fault index
