考虑多尺度储能需求的电氢储能系统优化配置：一种双层多步方法

Electric-hydrogen coupled systems (EHCSs) integrated with renewable energy offer significant advantages for providing clean energy provision yet face supply-demand imbalances across various timescales. This paper proposes a two-layer, multi-step optimal sizing framework for electric-hydrogen energy storage to address multi-scale energy storage requirements. The first step, the optimal sizing layer, determines the initial storage size and the number of clusters via genetic particle swarm optimization (GPSO). The second step, the operational layer, optimizes long-duration hydrogen storage on an annual scale to minimize imbalance risks and determine daily net hydrogen energy. The daily net hydrogen energy derived from the second step, along with the original renewable energy and load data, are subsequently processed using an optimization-based time series method combined with validation and supplementation techniques to generate short-duration typical scenarios. The final step involves optimizing electric-hydrogen energy storage based on the basis of these typical daily scenarios, and converts the results into annual costs for iterative refinement through GPSO. The optimization results demonstrate that the proposed approach achieves a 17.63% reduction in the supply-demand imbalance risk cost and lowers the daily total operation costs by 18.3% compared with the traditional method. Additionally, the adopted optimization-based time series scenario processing method resulted in a 14.43% reduction in the root mean square error of the clustering results and a 27.8% decrease in the total reconstruction error.