Data-driven robust stochastic optimization for power systems operations

Abstract The uncertainty of renewable energy brings significant challenges to the operations of power system with the increasing penetration of renewable energy, such as wind power generation and photovoltaic power generation. The traditional optimization methods are not suitable to solve the problems of the power system operations anymore. Robust optimization (RO) method and stochastic optimization (SO) method are usually to solve uncertain optimization problems. However, RO often suffers conservativeness because of using an uncertainty set to cover all possible values for the uncertainty variable, while the computation cost of SO is prohibitive as SO considers random variables follow a predefined probability distribution based on the historical data. To overcome these weaknesses, a distributionally robust optimization (DRO) model, as well as a robust stochastic optimization (RSO) method, is proposed by combining RO with SO. In this chapter, the applications of using RO, DRO, and RSO to solve the distribution network reconfiguration problems considering the uncertain connected renewable energy resources, ACOPF problems with uncertain wind power, the unit commitment problems considering the uncertainties causing by wind power connected, and the dynamic economic dispatch model considering the uncertainty of wind power are presented.