A Target-Driven Unit Commitment Model Based on Robust Uncertainty Satisficing Optimization

Motivated by commercial requirements for predictable operating cost budgets, this paper proposes a target-driven unit commitment (UC) model based on robust uncertainty satisficing optimization (RUSO) optimization approach. The UC problem is formulated via a two-stage framework, in which the UC decisions are obtained in a here-and-now manner, and the economic dispatch (ED) solutions are wait-and-see decisions, which are made after the realization of uncertain solar output limits. In RUSO, the operators can set their satisfactory cost target and obtain robust decisions to control the actual operating cost, where the realized data record of solar output is used. Based on linear decision rule (LDR) approximation, the ED variables are assumed as functions to be affinely adaptive to solar prediction error, and the distribution deviation of solar power is further incorporated to make the RUSO model flexible and tractable. With real-world data, we simulate solar penetrations into the IEEE 14-bus test system. The corresponding results illustrate that the RUSO model can express the level of conservativeness inherent in the budget target setting and manage the cost overrun risk well while achieving superior out-of-sample performance.