Early Warning Condition Transient Stability on South Sulawesi System using Extreme Learning Machine

Abstract

The electrical systems, the addition of loads can result in fewer stability limits, if there is interference, it can cause black out. In this study analyzing early warning, by observing the limits of stability in the event of a disturbance before black out in the South Sulawesi electricity system. This study observed an early warning system consisting of 44 buses and 15 generators using a Voltage stability margin (VSM) in the event of a disruption. From the training data about each disruption from various buses that occur then learning to use Extreme Learning (ELM) engines is used to detect early warnings during transient conditions. From the ELM simulation results can work quickly 0.0001 and 0.0024 and the error value is low so that it can be known before a blackout occurs.