An Extreme Scenario Reduction Algorithm for Multi-time Scale Net-Load

Due to the uncertainty of renewable energy power, the power system needs to apply various types of flexible resources to balance supply and demand on multiple time scales. For multi-timescale planning and scheduling problems, the construction of typical scenarios of source-load is inseparable. Extreme scenarios are particularly important for power systems. For this reason, an extreme scenario reduction algorithm based on multi-time scale net-load is proposed in this paper. The study proposes a multi-time scale decoupling method based on annual net-load, decomposing the time series into monthly, weekly, and daily scales, with the time scales representing different net-load features. A scenario reduction algorithm is proposed that can select typical scenarios while retaining extreme scenarios and assign corresponding probability values to the extreme scenarios. The algorithm example is analyzed using data from Zhejiang Province, China, for 2019 to illustrate the effectiveness of the proposed algorithm.