Date of Award


Degree Type


Degree Name

Doctor of Philosophy in Electrical Engineering


Electrical, Computer, and Biomedical Engineering

First Advisor

Haibo He


Microgrids are small-scale controllable power systems composed of distributedgeneration (DG) units, energy storage systems, and loads. Microgrids can be operated with the bulk power systems or independently from bulk power systems as an island. To stabilize frequency and voltage in islanded microgrids, a three-level hierarchical control strategy is widely implemented which includes primary control, secondary control, and tertiary control. This centralized control strategy suffers from a single point of failure and high communication cost. In addition, it cannot accommodate the plug-and-play operation of DGs in microgrids.

To overcome these challenges, this dissertation focuses on developing distributed robust and resilient control algorithms to stabilize frequency and voltage in microgrids, e.g, intermittent control, finite-time control, and robust control. Specifically, this dissertation first proposes a distributed secondary control protocol by using intermittent control technique. The proposed control method aims to restore frequency and voltage to their nominal values as well as maintain active power sharing in mean square. Second, this dissertation designs a distributed secondary controller which can restore frequency to the nominal value and achieve precise active power sharing in finite-time. Next, considering the effect of external disturbances and using an event-triggered strategy, this dissertation develops a robust distributed secondary control scheme under directed communication graphs. Then, this dissertation investigates the frequency control method in multi-microgrids. Next, this dissertation designs a distributed finite-time economic dispatch algorithm which is robust against denial-of-service (DoS) attacks. Finally, this dissertation proposes a deep reinforcement learning based phase-locked loop (PLL) structure for microgrids with fast response speed. Numerous simulations are conducted to validate the effectiveness of the proposed methods.

Available for download on Thursday, September 05, 2024