Identification of Dynamic Equivalents for MicroGrids with High Penetration of Solar Energy using Artificial Neural Networks

Abstract

Dissemination of small size dispersed microgeneration, connected to Low Voltage (LV) distribution systems is expected to become an effective mean to face the continuous demand growth. Dynamic behaviour analysis is required in Medium Voltage (MV) distribution grids with dispersed generation if islanding operation is allowed. When large scale integration of Photovoltaics (PV) in LV grids occurs a special care is needed when performing these simulations. This paper describes an approach based on nonlinear modelling and identification of dynamic systems using Artificial Neural Networks (ANN) used to build dynamic equivalents for Micro Grids (MG), which can be integrated in dynamic simulation tools. A test system with high penetration of PV is used to evaluate the performance of this approach.