The objective of this project is to discover robust and cost efficient power generation scheduling, in order to maintain a smart and secure power grid while ensuring high utilization of renewable energy. The approach is to study innovative robust and chance constrained optimization models along with solution methodologies to provide thermal unit commitment decisions incorporating intermittent renewable generation. Intellectual Merit: This project proposes one of the first studies on two-stage robust and chance-constrained optimization methods to solve power grid optimization problems. The proposed project will enrich scientific methods to solve power grid optimization problems under renewable generation uncertainty, while ensuring system reliability. In addition, the proposed solution approach will lead to methodology innovations for robust and chance constrained optimization, including deriving exact separation algorithms for two-stage robust unit commitment problems and proving the convergence of the sample average approximation algorithm for two-stage chance constrained stochastic programs. Broader Impacts: The proposed research is extremely beneficial for the reliability runs for system operators to achieve cost efficiency and security. It will also provide a tool to estimate the storage capacity requirement in order to guarantee a certain percentage (e.g., 80%) of electricity usage coming from renewable. In addition, the collaboration with national research labs facilitates testing real data and implementation at energy markets in short time, which will immediately benefit the society. Finally, results from this project will be disseminated through multiple means, including journal publications and conference presentations, and underrepresented students will be involved in all aspects of this research effort.
