Eur. Phys. J. AP
Volume 16, Number 3, December 2001
|Page(s)||195 - 208|
|Section||Nanomaterials and Nanotechnologies|
|Published online||15 December 2001|
Fuzzy logic controllers for electrotechnical devices - On-site tuning approach
Institut National Polytechnique de Toulouse,
École Nationale Supérieure d'Électrotechnique, d'Électronique,
d'Informatique et d'Hydraulique de Toulouse, Laboratoire d'Électrotechnique et
d'Électronique Industrielle (UMR CNRS n° 5828), 2 rue Camichel, BP 7122,
31071 Toulouse Cedex 7, France
2 Institut National Polytechnique de Toulouse, École Nationale Supérieure d'Électrotechnique, d'Électronique, d'Informatique et d'Hydraulique de Toulouse, Laboratoire d'Électronique, Électrotechnique et Systèmes, Unité mixte de recherche UTBM & UFC associée à l'INRETS LRE T31 L2ES - IGE Parc Technologique, 2 avenue Jean Moulin, 90000 Belfort, France
Corresponding author: email@example.com
Revised: 8 March 2001
Accepted: 20 July 2001
Published online: 15 December 2001
Fuzzy logic offers nowadays an interesting alternative to the designers of non linear control laws for electrical or electromechanical systems. However, due to the huge number of tuning parameters, this kind of control is only used in a few industrial applications. This paper proposes a new, very simple, on-site tuning strategy for a PID-like fuzzy logic controller. Thanks to the experimental designs methodology, we will propose sets of optimized pre-established settings for this kind of fuzzy controllers. The proposed parameters are only depending on one on-site open-loop identification test. In this way, this on-site tuning methodology has to be compared to the Ziegler-Nichols one's for conventional controllers. Experimental results (on a permanent magnets synchronous motor and on a DC/DC converter) will underline all the efficiency of this tuning methodology. Finally, the field of validity of the proposed pre-established settings will be given.
PACS: 84.60.Bk – Performance characteristics of energy conversion systems; figure of merit / 07.05.Fb – Design of experiments / 07.05.Mh – Neural networks, fuzzy logic, artificial intelligence
© EDP Sciences, 2001
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.