ISE Graduate Seminar
Exploiting Almost Symmetries in the Unit Commitment Problem
Unit commitment (UC) is the problem of minimizing the production cost of a set of power generators so that demand can be satisfied. It is often the case that generators in a real-world UC instance have identical operating constraints. Recent work has shown that significant computational advantage can be achieved by using symmetry exploiting techniques to handle identical generators. However, generation providers do not always offer consistent bids with such generators. This hides the symmetry in the problem and can prevent solvers from taking advantage of this structure. This work seeks to address this problem by creating and solving a symmetric relaxation of the UC problem by modifying the cost coefficients of nearly identical generators. We investigate the benefits of using a symmetric relaxation and the trade-offs between solution speed and quality when costs differ.
Jonathan Schrock is a PhD student in the Industrial and Systems Engineering department at the University of Tennessee Knoxville. He studied mathematics at the University of Tennessee and received his Master of Science degree in 2015 focusing on symmetry detection in linear programs. During this time, he worked on projects in quantum computing, graph generation, and HPC benchmarking at Oak Ridge National Laboratory. Currently, Jonathan is working on the unit commitment problem using symmetry methods to improve solution time and quality and modeling pumped storage hydropower systems.
Friday, October 29, 2021 at 3:30pm to 4:30pmVirtual Event