Unified Stochastic and Robust Unit Commitment

Chaoyue Zhao

doi:10.1109/TPWRS.2013.2251916

Return

Unified Stochastic and Robust Unit Commitment

2013-08-01

317

PRE

Chaoyue Zhao *

Yongpei Guan

DOI：10.1109/TPWRS.2013.2251916

Original

Original request for help

Save

Abstract

En 中文

Due to increasing penetration of intermittent renewable energy and introduction of demand response programs, uncertainties occur in both supply and demand sides in real time for the current power grid system. To address these uncertainties, most ISOs/RTOs perform reliability unit commitment runs after the day-ahead financial market to ensure sufficient generation capacity available in real time to accommodate uncertainties. Two-stage stochastic unit commitment and robust unit commitment formulations have been introduced and studied recently to provide day-ahead unit commitment decisions. However, both approaches have limitations: 1) computational challenges due to the large scenario size for the stochastic optimization approach and 2) conservativeness for the robust optimization approach. In this paper, we propose a novel unified stochastic and robust unit commitment model that takes advantage of both stochastic and robust optimization approaches, that is, this innovative model can achieve a low expected total cost while ensuring the system robustness. By introducing weights for the components for the stochastic and robust parts in the objective function, system operators can adjust the weights based on their preferences. Finally, a Benders' decomposition algorithm is developed to solve the model efficiently. The computational results indicate that this approach provides a more robust and computationally trackable framework as compared with the stochastic optimization approach and amore cost-effective unit commitment decision as compared with the robust optimization approach.

Keywords:

Benders' decomposition

mixed-integer linear programming (MILP)

robust optimization

stochastic optimization

unit commitment

AI Summary

Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.