A Dual-Layer Optimization Model for Off-Grid Wind-Solar-Storage PEM Electrolyzer Hydrogen Production Based on NSGA-II: Scenario Generation Combining Kernel Density Estimation and Copula

This paper investigates an off-grid wind-solar-storage PEM electrolysis hydrogen production system, proposing a capacity optimization configuration model for such systems. First, a combined method of kernel density estimation and Copula functions is employed to construct the joint distribution of wind and solar power outputs. Subsequently, K-means clustering is applied for dimensionality reduction, generating scenarios that balance correlation and representativeness for wind and solar resources. Second, addressing system economics and stability, a two-layer capacity optimization model for the off-grid wind-solar-storage PEM electrolysis hydrogen production system is established based on the NSGA-II algorithm. Finally, a specific case study is selected to validate the proposed two-layer optimization model.