Amid the global push for renewable energy integration, power systems are increasingly facing uncertainties due to the variable nature of renewable energy sources (RES) and demand response (DR). A recent study published in Engineering offers a novel framework to address these challenges through robust optimization techniques, specifically focusing on decision-dependent uncertainties (DDUs) in power systems with high shares of renewables.

The research, led by Feng Liu, Yunfan Zhang from Tsinghua University, and Yifan Su from Chongqing University, explores the complexities of DDUs, which arise when uncertainties in power system dispatch are influenced by operational decisions. This is in contrast to traditional decision-independent uncertainties (DIUs), where uncertainties are exogenous and unaffected by decisions. The study systematically investigates the characteristics of DDUs and proposes a robust optimization framework to handle these uncertainties effectively.

The authors introduce the concept of "separability" to describe the dependency structure of DDUs, demonstrating that any DDU set can be decomposed into a decision-independent support set and a coupling function that describes the dependency on decisions. This finding bridges the gap between DDUs and conventional DIUs, suggesting that existing algorithms for DIUs can be extended to address DDUs.

The study highlights that DDUs introduce new characteristics to power system dispatch problems, such as non-convexity, which complicates the optimization process. However, the authors argue that DDUs also provide an opportunity for bilateral matching between uncertainties and system flexibility, enhancing the overall robustness of power system operations. They propose an improved solution algorithm that adapts to the convexity or non-convexity of the problem, ensuring robust feasibility and optimality.

The research provides several illustrative applications to demonstrate the significance of considering DDUs in power system dispatch. For instance, in the context of wind power, the study shows that modeling wind generation output as a DDU rather than a DIU can significantly improve frequency regulation performance. Similarly, in demand response programs, accounting for DDUs can enhance the robustness of dispatch decisions, ensuring reliable power system operations.

The study concludes that while DDUs introduce complexities, they also offer opportunities for more flexible and robust power system dispatch. The proposed framework and algorithms provide a systematic approach to handling DDUs, laying a foundation for the reliable operation of power systems dominated by renewable energy sources. This work underscores the importance of considering DDUs in power system planning and operations, paving the way for future research and practical applications in the field of smart grid technologies.

The paper “On Decision-Dependent Uncertainties in Power Systems with High-Share Renewables,” is authored by Yunfan Zhang, Yifan Su, Feng Liu. Full text of the open access paper: https://doi.org/10.1016/j.eng.2025.07.013. For more information about Engineering, visit the website at https://www.sciencedirect.com/journal/engineering.