Subgrid Identification and Islanding in Future Medium Voltage Smart Grids based on Container Virtualisation and Machine Learning

Abstract:
The paper proposes a strategy for the adaptive identification, determination, and islanding of subgrids in medium voltage (MV) smart grids. In contrast to the intended islanding of grids, the term subgrid describes a subordinate energy system of undetermined size which is physically isolated from the interconnected grid and continues to operate after disconnection. This process takes place outside of the control of the distribution system operator (DSO). Due to risks for people and equipment, subgrids must be identified reliably and quickly in order to disconnect them from the power supply or transfer them to an island operation. A Subgrid Identification and Islanding Algorithm (SIS) is currently being implemented and integrated into practical application in the publicly funded research project i-Autonomous at the Technical University of Dortmund in a Hardware-in-the-Loop (HIL) laboratory setup. Additionally, the implementation is done by a novel engineering process based on a virtualisation of smart grid services using the standard IEC 61850, a modular integration of functions on hardware devices using the technologies Docker and Kubernetes and four selected algorithms based on different methologies.