Probabilistic Analysis of Time-Series Production Simulation for Large-Scale Renewable Energy Bases in Desert-Gobi-Wasteland Regions Based on Dimension-Adaptive Sparse Grid Interpolation
Large-scale renewable energy bases in desert-gobi-wasteland regions, typically connected to load centers via long-distance weak tie-lines and high-voltage direct current (HVDC) corridors, face significant challenges in accurately and efficiently evaluating renewable energy accommodation rates and transmission corridor utilization. This study addresses the computational inefficiency of probabilistic time-series production simulation under dual uncertainties—stochastic renewable generation and planning variables such as installed capacity and peak load. A dimension-adaptive sparse grid interpolation (DASGI) surrogate model is proposed to approximate the complex original time-series production simulation model. The method integrates Monte Carlo sampling with the surrogate model to enable rapid probabilistic analysis and risk assessment. Experimental results demonstrate that the DASGI surrogate model achieves high fidelity with the original model while substantially reducing computation time. Furthermore, incorporating additional configuration points enhances the model's ability to precisely identify potential violation risks. The proposed approach offers a computationally efficient tool for uncertainty quantification in the planning of large-scale renewable energy bases and their HVDC transmission corridors, providing theoretical support for coordinated planning studies.