Title: ANN based real time incipient fault detection and protection system for induction motor

Authors: Makarand S. Ballal, Hiralal M. Suryawanshi, Mahesh K. Mishra

Addresses: Office of the Dy. Executive Engineer, Testing Unit, Deepnagar Substation, Bhusawal, 425201 India. ' Electrical Engineering Department, Visvesvaraya National Institute of Technology, Nagpur 440011, India. ' Electrical Engineering Department, Indian Institute of Technology, Chennai 600036, India

Abstract: Artificial neural network (ANN) has its unique advantage in the area of incipient faults detection. This article presents an ANN based real time fault detection and protection system for two types of incipient faults viz. inter-turn insulation failure and bearing wear in single-phase induction motor. The ANN fault detection (ANNFD) program is developed in C++ and implemented using a PC based DSP controller board. The ANN is trained and tested by collecting the online experimental data for five input parameters viz. motor intake current, rotor speed, winding temperature, bearing temperature and noise. The results are compared with the signature of measurable parameters. The results of evaluation indicate that the system produces satisfactory performance for the fault detection as well as for the protection of motor.

Keywords: induction motors; incipient faults; condition monitoring; motor protection; insulation failure; bearing wear; artificial neural networks; ANNS; fault detection; ANNFD; motor intake current; rotor speed; winding temperature; bearing temperature; noise.

DOI: 10.1504/IJPEC.2009.027940

International Journal of Power and Energy Conversion, 2009 Vol.1 No.2/3, pp.125 - 142

Received: 29 May 2008
Accepted: 09 Nov 2008
Published online: 20 Aug 2009 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article