Speed estimation using neural network in vector controlled induction motor drive

  • Authors:

  • F. Haghgoeian;M. A. Ouhrouche;J. S. Thongam

  • Affiliations:

  • Electric Machines Identification and Control Laboratory, Department of Applied Sciences, University of Quebec at Chicoutimi, Chicoutimi, QC, Canada;Electric Machines Identification and Control Laboratory, Department of Applied Sciences, University of Quebec at Chicoutimi, Chicoutimi, QC, Canada;Electric Machines Identification and Control Laboratory, Department of Applied Sciences, University of Quebec at Chicoutimi, Chicoutimi, QC, Canada

  • Venue:
  • CONTROL'05 Proceedings of the 2005 WSEAS international conference on Dynamical systems and control
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper presents a speed estimation method using neural networks (NN) in a vector controlled (VC) induction motor drive. The estimation algorithm is implemented using a Jordan recurrent NN structure where training of the NN is done online using back-propagation algorithm. Two back emf models are used in order to realize the reference and the adaptive models from which depending upon the speed error back emf error is generated which is used for training the NN. Results of real-time digital simulation using RT-Lab show good estimation accuracy. This achievement is believed to be an important contribution to sensorless vector control of induction motor drive.