Data-driven Optimization Approaches to the Power System Planning under Uncertainty

TL;DR

This paper reviews and compares three key optimization methods—stochastic programming, robust optimization, and distributionally robust optimization—for power system planning under uncertainty, emphasizing their frameworks, strengths, and future prospects.

Contribution

It introduces and compares three major optimization techniques under uncertainty, including a novel distributionally robust approach, for power system planning.

Findings

DRO combines strengths of SP and RO.

Each method has unique advantages and limitations.

Future research directions are discussed.

Abstract

In order to protect the environment and address fossil fuel scarcity, renewable energy is increasingly used for power generation. However, due to the uncertainties it brings to electricity production, deterministic optimization is no longer sufficient for operational needs. Therefore, a large number of optimization techniques under uncertainty have been proposed, which provide good ways to address uncertainties. This paper selects three of the more important optimization techniques under uncertainty to introduce: stochastic programming (SP), robust optimization (RO), and a novel approach named distributionally robust optimization (DRO) based on the first two. We explain the basic framework and general process of each approach using specific examples. The focus is on how each method addresses the uncertainties. In addition, we also compare their strengths and weaknesses and discuss future research directions.