A New Algorithm for Automatic Generation of Fuzzy Rules. Application to power systems controllers

 

A.M.Luciano, D.Lauria, Ettore Napoli

abstract

The present paper presents a new Fuzzy algorithm for the identification of the input-output characteristic of a generic system.

The method allows to obtain a simple analytical form for the identified function.

The identification is led to a non linear optimization problem that we approach by means of the Downhill Simplex or Quasi-Newton method.

An application is performed with reference to the electrical power systems stability improvements. The method for the generation of fuzzy rules is applied to a fuzzy logic variable structure FACTS controller. Our approach effectiveness is verified through numerical applications.


Back