Optimization and Model-predictive Control for Overload Mitigation in Resilient Power Systems

Abstract: The National Academy of Engineering named the electric power grid the greatest engineering achievement of the 20th century. However, as recent large-scale power grid failures illustrate, the (electro-mechanical) electric grid is being operated closer and closer to its limits. Due to the protracted and cost-intensive nature of upgrading energy infrastructures, major research initiatives are now underway to improve the utility of the existing infrastructure. One important topic is contingency management. Accordingly, this dissertation comprises of practical, yet rigorously justified, feedback control algorithms that are suitable for power system contingency management. The main goals of the algorithms are to prevent or mitigate overloads on network elements (e.g. transmission lines and transformers).
The ideas of balancing economic and safety criteria are developed and implemented with a receding-horizon model-predictive controller (RHMPC) for electric transmission systems with energy storage and renewables. The novel RHMPC scheme employs a lossy "DC" power flow model and is proven to alleviate conductor temperature overloads and returns the system to an economically optimal state. The algorithms proposed herein are, essentially, "closing the loop" in contingency management, and will help bring the electric grid into the 21st century.