High-fidelity modeling and resilience optimization of microgrids

Microgrid is the smart renewable energy system for sustainable power supply of critical systems (e.g., buildings, hospital, transport, communication, data center, defence). The vulnerable components and sophisticated cyber-physical network bring great threats to microgrid resilience. A high-fidelity model is the key to achieve the guaranteed resilience of microgrid.

But the complex system with nonlinearities and time-varying characteristics as well as cyber-physical vulnerabilities hinders the high-fidelity modeling.

This project will develop a high-fidelity microgrid digital twin model based on physics analysis and data-driven learning algorithms.

It consists of a self-evolving modelling approach to reveal the inner information of the physical system in real time, and a self-detection based model updating approach to guarantee the data quality under cyberattacks.

Next, based on the high-fidelity digital twin, an integrated resilient management scheme will be developed for health monitoring, anomaly detection and overall optimization. A microgrid digital twin will be built for a real-world microgrid hardware system as a verification and demonstration.

The proposed work will provide fundamental research for digital twin development and resilience optimization of smart microgrids, as well as other similar cyber-physical systems. It will achieve highly reliable and sustainable power supply for the smart and sustainable society.