Date

2018

Document Type

Dissertation

Degree

Doctor of Philosophy

Department

Electrical Engineering

First Adviser

Liu, Wenxin

Abstract

Cyber system is a fairly important component of the energy systems. The network imperfections can significantly reduce the control performance if not be properly treated together with the physical system during the control designs. In the proposed research, the advanced controls of cyber-physical energy systems are explored in depth. The focus of our research is on two typical energy systems including the large-scale smart grid (e.g. wide-area power system) and the smart microgrid (e.g. shipboard power system and inverter-interfaced AC/DC microgrid). In order to proactively reduce the computation and communication burden of the wide-area power systems (WAPSs), an event/self-triggered control method is developed. Besides, a reinforcement learning method is designed to counteract the unavoidable network imperfections of WAPSs such as communication delay and packet dropout with unknown system dynamics. For smart microgrids, various advanced control techniques, e.g., output constrained control, consensus-based control, neuro network and game theory etc., have been successfully applied to improve their physical performance. The proposed control algorithms have been tested through extensive simulations including the real-time simulation, the power-hardware-in-the-loop simulation and on the hardware testbed. Based on the existing work, further research of microgrids will be conducted to develop the improved control algorithms with cyber uncertainties.

Share

COinS