Two-stage Hybrid Optimization of Aggregated Distributed Generalized Energy Storages for Complete Uncertainty Elimination

The intrinsic uncertainties in the widespread distributed renewable energy resources pose considerable threats to the secure and reliable operation of distribution networks (DNs). To fully absorb the uncertainties in DN, this paper proposes a novel two-stage hybrid optimization approach for the distributed generalized energy storage systems (DGESSs) by integrating the day-ahead optimal scheduling with the realtime uncertainty mitigation. First, considering the features of a large population of DGESSs, an inner approximation-based aggregation model is proposed to effectively aggregate various DGESSs into an equivalent energy storage with the identical form. Then, the optimal scheduling of the aggregated energy storage systems (AESSs) is cast as a two-stage hybrid model combining stochastic programming and robust optimization to optimize of day-ahead scheduling baseline and the real-time response rules. Consequently, the real-time power adjustments of AESSs can be on-line determined according to the pre-optimized affine rules. Furthermore, the originally intractable hybrid model is converted into a solvable form with the minimum information of the uncertainties. Finally, numerical tests on the modified IEEE 123-bus distribution system validate the effectiveness of the proposed approach in mitigating the impact of uncertainties on the upstream main grid, improving the voltage quality, and enhancing the economic efficiency of DN.