Optimized instantaneous torque control of switched reluctance motor by neural network Conference Paper

abstract

• An artificial neural network (ANN) based torque control scheme of switched reluctance motor (SRM) was used which generates optimal current profiles to reduce torque pulsation. For each speed and torque below base speed, several current profiles were observed to produce desired torque without any pulsation. Selected optimal current profiles were used to train a neural network which recreates the optimal profile on-line. Torque pulsation can be reduced by selecting, by dynamic model, current profiles which have a wider turn on and turn off characteristics, thus allowing ANN have a better chance to learn the profiles. The experimental results agreed well between commanded and developed torque validating the effectiveness of the torque control scheme.

name of conference

• IAS '97. Conference Record of the 1997 IEEE Industry Applications Conference Thirty-Second IAS Annual Meeting

authors

• Ehsani, Mehrdad

published proceedings

• IAS '97 - CONFERENCE RECORD OF THE 1997 IEEE INDUSTRY APPLICATIONS CONFERENCE / THIRTY-SECOND IAS ANNUAL MEETING, VOLS 1-3

author list (cited authors)

• Rahman, K. M., Rajarathnam, A. V., & Ehsani, M.

citation count

• 41

complete list of authors

• Rahman, KM||Rajarathnam, AV||Ehsani, M

publication date

• January 1997

publisher

• Institute of Electrical and Electronics Engineers (IEEE)  Publisher

published in

• Conference Record of the IEEE Industry Applications Conference  Journal

Digital Object Identifier (DOI)

• 10.1109/ias.1997.643123

International Standard Book Number (ISBN) 10

• 0-7803-4067-1

start page

• 556

end page

• 563

volume

• 1

URL

• http%3A%2F%2Fdx.doi.org%2F10.1109%2Fias.1997.643123