Adaptive Neuro-Fuzzy Synchronization in Isolated Power Systems with High Wind Penetration

Abstract

Isolated power systems (IPSs) worldwide are traditionally powered by diesel generators that are very expensive to run and produce harmful emissions. In order to mitigate these problems, wind turbines are being introduced into existing IPSs. Although this integration has been reasonably effective at reducing running costs and emissions, high levels of wind penetration cause large system frequency variations, resulting in a prolonged synchronization process for newly dispatched diesel generators. Long synchronization can compromise the stability of a small IPS. This paper examines the diesel synchronization problem using a real IPS as a case study and offers a solution by introducing the concept of predictive synchronization based on adaptive neuro-fuzzy systems. Simulation results demonstrate a significant reduction in diesel generator synchronization times.