Determinación de los parámetros para mejorar el control automático de generación en sistemas eléctricos de potencia con interconexiones

Abstract

This document develops a method to find the appropriate parameters to perform automatic generation control (AGC), focusing on the analysis of the frequency bias factor β necessary for determining the area control error (ACE), in order to reduce the response of the secondary frequency regulation (SFR) to any eventuality that the Ecuadorian electrical system may present. The method is premised on estimation and supervised machine learning based on historical records of frequency and power events that allows obtaining the necessary information to perform the AGC. The national interconnected system (SNI) is modeled under different conditions, normal and atypical. The bias values obtained were simulated and verified under distinct scenarios in the DigSilent PowerFactory computer program.