This project involves conducting data analytics and designing optimization algorithms to develop and solve a resource allocation problem in the power systems domain. Its deliverables will improve the efficiency and equitability of the process for restoring power in affected areas of an electric power network. We use power systems data to develop optimization models and metrics for a trade-off between efficiency and fairness in power restoration decision-making. While efficiency (e.g., minimizing overall energy loss) is essential, incorporating fairness ensures an equitable distribution of resources across all impacted regions. In this remote project, the SUDS scholar will work under the guidance of data science and optimization faculty experts to complete a series of weekly research tasks.

After receiving some training on relevant machine learning and optimization models, the SUDS Scholar will be assigned weekly tasks that may include data analytics, formulating optimization models, implementing machine learning and Gurobi optimization models, conducting computational and data-intensive experiments on real and/or synthetic datasets, and developing exact or heuristic optimization methods to improve solution efficiency and accuracy. The expected outcome of this research is contributing reliable, open-source, and reproducible models and algorithms to the broader research communities in power systems optimization and data science.