Academic Journals Database
Disseminating quality controlled scientific knowledge

Condition Monitoring and Fault Diagnosis of Serial Wound Starter Motor with Learning Vector Quantization Network

Author(s): R. Bayir

Journal: Journal of Applied Sciences
ISSN 1812-5654

Volume: 8;
Issue: 18;
Start page: 3148;
Date: 2008;
VIEW PDF   PDF DOWNLOAD PDF   Download PDF Original page

Keywords: Starter motor | condition monitoring | fault diagnosis | learning vector quantization neural network

In this study, a Graphical User Interface (GUI) software for real time condition monitoring and fault diagnosis of serial wound starter motors has been developed using Learning Vector Quantization (LVQ) neural network. The starter motors are serial wound dc motors which enable the Internal Combustion Engine (ICE) to run. When the starter motor fault occurs, the ICE cannot be run. Therefore, condition monitoring and pre-diagnosis of starter motor faults are important. The information of voltages and currents is acquired from the starter motor via data acquisition card and transferred to the program. With this program using LVQ network, six faults observed in the starter motors were successfully detected and diagnosed in real time. The GUI software makes it possible to condition monitoring and diagnose the faults in starter motors before they occur by keeping fault records of the past occurrences. This system can be used in service shops and in test departments of starter motor manufacturers. In addition, this system has potential to be used for real time condition monitoring and fault diagnosis of vehicles with the help of industrial computers.
Affiliate Program     

Tango Jona
Tangokurs Rapperswil-Jona