AMPS: Novel Combinatorial Optimization Techniques for Smartgrids and Power Networks

William Marsh Rice University, Houston TX

Investigators

Abstract

Climate change, safety concerns, obsolescence of utility structures and equipment, and the goal of incorporating a higher level of renewable generation are now placing the traditional power grid and the assumption of easy access to it at risk. Envision the development of a new generation of small, autonomous networks of storage (batteries), solar panels, and even small turbines or engines to provide an alternative power source to small communities – as small as a single building or group of buildings. Such grids, called microgrids, could flexibly interconnect or remain independent, according to the economics of physical demands. The design and operation of such systems will require novel technical developments that are explored through this project. In particular, this project will develop new theory and computationally efficient algorithms that address these computational challenges related to microgrids. The project will advance the knowledge base in microgrids and their utility within the electrical grid structure. Full translation and accessibility by the burgeoning smart grid community will allow rapid adoption. Regarding education, the proposed techniques and other research at the intersection of discrete optimization and power systems offer a rich opportunity to provide a graduate course on this particular topic. The proposed project will also allow the PI to continue ongoing efforts to increase the engagement of underrepresented groups within STEM and create new opportunities. New technological challenges also come with the rise of smart microgrids in the electrical power system. In particular, this project utilizes polyhedral theory, branch decompositions, and other combinatorial optimization techniques, advancing the knowledge base in combinatorial optimization while addressing computational challenges related to battery storage, electrical grid partitioning and observability, and power flow. We envision that these techniques will be practical and scalable. This research's theoretical and computational results will be disseminated through research publications, conferences, and software. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.