Eur. Phys. J. AP
Volume 6, Number 1, April 1999
|Page(s)||61 - 70|
|Published online||15 April 1999|
On the robust control of an induction machine: A complete design and realization
CEGELY (CNRS No 5005), École Centrale de Lyon, B.P. 163, 69131
Écully Cedex, France
Corresponding author: email@example.com
Revised: 19 October 1998
Accepted: 23 December 1998
Published online: 15 April 1999
Electrical traction of vehicle needs accurate control of torque and flux. DC machines are Maynly used but they are expensive and need heavy Mayntenance. Since some years, the use of field oriented vector control induction machines allows to fulfill the same objectives with lower cost. Unfortunately, they may present a decrease of their performances and an instability when their parameters vary with the temperature or the magnetic state and in the presence of converter or measurement noises. Therefore it is necessary to design effective control laws, which are especially less sensitive to these perturbations and variations. In this work, we have undertaken the conception and the realization of a voltage vector flux control whose inputs have been uncoupled by compensating terms. The torque and the flux regulations are realized at two different levels: an internal regulation loop for stator currents, and an external one for flux and torque. In order to achieve robust stability and performance objectives, we have involved several new methods in the doMayn such as H∞ control designed by genetic algorithms and reduced order extended Kalman filtering in the synchronous frame. Simulations will show the method efficiency and experimental results validate the control strategy.
PACS: 07.05.Dz – Vacuum gauges / 84.50.+d – Electric motors
© EDP Sciences, 1999
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.