A Robust Stochastic Optimization for Locally Ideal Unit Commitment Considering Renewable Energy Sources

Abstract

The unit commitment (UC) is a fundamental task in day-ahead electricity market. However, the proliferation of renewable energy sources (RES), especially wind power and solar power, significantly influences the economics of UC, due to their inherent uncertainties. Therefore, this paper presents a general robust stochastic optimization (RSO) framework for performing UC with involving the uncertainties of RES. Firstly, a typical UC model is presented, which determines the on/off state, base-point generation, and reserve level for units. And then, the intractable quadratic objective is linearized using a locally ideal piecewise formulation, so that a tractable mixed integer linear programming (MILP) based UC model is obtained. Furthermore, an event-wise RSO framework is employed to deal with the uncertainties, in which the uncertain RESs are captured by an event-wise ambiguity set. Moreover, a practical way is designed to apply the RSO framework on the presented UC model. Finally, using realworld RES data from Belgian Transmission System Operators, experimental results on IEEE RTS 24-bus system demonstrate the effectiveness of the presented method.