سال انتشار: ۱۳۷۵

محل انتشار: یازدهمین کنفرانس بین المللی برق

تعداد صفحات: ۱

نویسنده(ها):

Menaj – Amirkabir University IRAN
Seifipour – Amirkabir University IRAN
Abedj – Amirkabir University IRAN

چکیده:

It is well-known that in order to apply fuzzy inference models as linguistic controllers in any control system, we need to know some parameters such as number of fuzzy partitions, shape and specifications of the membership functions.
Generally speaking, these parameters are chosen intuitively. In this paper we show that in compared with intuitive selection of these parameters the performance of fuzzy control systems will be improved by using genetic algorithms as well as Marquardt learning algorithm which is a modification of standard back propagation algorithm [6]. These two approaches ( genetic algorithms and back propagation training algorithm), yield appropriate parameters for fuzzy logic control system. In order to show the effectiveness of the proposed modifications on FLC, we apply them on a synchronous generator.