PERPERFORMANCE-PARAMETER PREDICTION OF DOUBLY FED INDUCTION GENERATOR (DFIG) USED IN WIND GENERATOR APPLICATION BY ARTIFICIAL NEURAL NETWORK (ANN)

Abstract

In modern large wind turbines, doubly-fed induction generators are commonly used. In these generators, both stator and rotor windings are connected to grid through power electronic DC/AC converters. The torque depends on the strength of two fields and the phase angle between them. The speed range is approximately + - 30% of the grid frequency. Considering these special features, the DFIG analysis is performed using ‘d q o’ transformation [1]. A design data bank is generated for machine ratings from 1000kW to 3000kW. With the help of Artificial Neural Networks (ANN), equivalent circuit parameters, stator and rotor voltage and current, and reactive power supplied to and received by the grid are predicted.