Evolutionary Algorithm EPSO Helping Doubly-Fed Induction Generators in Ride-Through-Fault

Abstract

A tuning process of the PI(Proportional-Integral)controller gains of a doubly-fed induction generator's (DFIG) rotor side converter is described in this work. The purpose is to tune PI controllers to help the DFIG to survive to network faults, avoiding being tripped-off by under-voltage relays.

The ride-through-fault capability of DFIGs improves system reliability and allows them to participate in the control and stabilization of a power system following system disturbances. The robust tuning of the DFIG rotor side converter's PI controllers may avoid undesired disconnections from the grid by,for instance, preventing over-currents in its variable frequency AC/DC/AC converter. This work presents an Evolutionary Particle Swarm Optimization-based (EPSO) approach to this tuning, with the aim of helping to limit the line-to-line voltage dip at the DFIG's terminals after a short-circuit, in order to avoid its tripping-off. The EPSO-based algorithm developed is validated at a typical Portuguese 15 kV Distribution Network with the integration of a DFIG, using the transient electromagnetic software package PSCAD/EMTDC.