Mean robust unit commitment for wind-storage systems considering wind and building load uncertainties

The increasing coupling between renewable generation and building energy consumption introduces composite uncertainties to power system operations. This paper proposes a Mean Robust Unit Commitment method for wind-storage integrated power grids, specifically addressing the aggregated forecast deviations of wind power and building loads. A unit commitment model is established, incorporating the start-up/shut-down characteristics of thermal units and the energy limits of storage systems. To balance robustness and computational efficiency, a Mean Robust Optimization framework is employed. By compressing historical error data into representative centroids within a Wasserstein ambiguity set, the chance constraints for reserve adequacy are reformulated into a tractable mixed-integer linear programming problem. Case studies on the IEEE 24-bus system validate that the proposed method effectively manages the net load volatility caused by wind and buildings while significantly reducing solution time compared to conventional distributionally robust optimization.