Eur. Phys. J. AP
Volume 4, Number 1, October 1998
|Page(s)||53 - 64|
|Published online||15 October 1998|
A few discussion on expert system development for electrical power systems-optimization of an inverter-motor of a railway traction chain
Laboratoire d'Électrotechnique et d'Électronique
Industrielle (UMR-CNRS 5828), I.N.P. Toulouse, E.N.S.E.E.I.H.T.,
2 rue Charles Camichel, 31071 Toulouse Cedex, France
2 Alstom Transport SA, rue du Docteur Guinier BP 4, 65600 Séméac, France
Accepted: 3 April 1998
Published online: 15 October 1998
This paper deals with the automatic design of static converters and electrical machines. Facing the multitude of such devices (converters–machines), the choice of the appropriate converter or machine is a delicate problem. A high level of expertise is required to determine the best structure for the considered application. This choice could be rapidly made if one had a system which incorporates a great knowledge of converters, machines and control methods. Firstly, the authors review the realized studies, then they propose this methodology for a concrete case in order to optimize an inverter–motor of a railway traction chain. Constraints are described in terms of input and output performances (losses, current, ...). These data are analyzed by using design rules and methods which allow the decrease of the manufacture cost of the inverter–motor. Rules and methods are implemented on a workstation by means of an “expert system shell” Smeci. The article proposes several tracks which involve organization, methodology, confirmation and implementation of a tool supporting a method. This process of implementation is the most efficient to fit into a constant process of improvement which is supported by the return on experience and by the evolution of the state of the art.
PACS: 02.60.x – Numerical approximation and analysis / 02.90.+p – Other topics in mathematical methods in physics / 01.50.Kw – Testing theory and techniques
© EDP Sciences, 1998
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.