Closing the Feasibility Gap: Modeling to Generate Alternatives for AC Topology Optimization

Abstract

Routine changes to power transmission systems' topology can alleviate congestion, maintain supply security, and increase utilization of renewable energy. Finding the optimal transmission network topology takes the form of a mixed-integer non-linear programming problem (MINLP), which can't presently be solved at scale in the required time. Various linearization methods exist, most famously the DC optimal transmission switching (DC-OTS) problem. Although this approximation sometimes yields good results, the resulting topology can often have no feasible solution to the non-linear problem. This paper introduces the use of modeling to generate alternatives (MGA) as an extension to the DC-OTS problem. We demonstrate that MGA successfully identifies good solutions for all 13 considered scenarios, all of which had infeasible topologies using DC-OTS.