A New Islanding Detection Method Using Transfer Learning Technique

Abstract

The increasing need for energy in the recent times is unprecedented, which is driving the penetration of renewable sources in distribution system in a big way. The increasing number of renewable sources in a system has made the operation, control and protection of the system very complex. One of the key issues in seamless interconnection of renewable energy sources to a system is islanding. This paper proposes a new method to detect islanding in an efficient way by employing transfer learning based technique for image classification. The results show that the proposed method can successfully classify islanding events with a good accuracy. � 2018 IEEE.