A Data-Driven Dynamic Temporal Correlation Modeling Framework for Renewable Energy Scenario Generation

The temporal correlation of renewable energy generation is inherently influenced by atmospheric systems. To capture its time-varying characteristics, a data-driven modeling framework is proposed for short-term renewable energy scenario generation. The framework uses a decoupled mapping approach to separately model marginal distributions and the correlation structure, with accuracy ensured through optimization based on proper scoring rules. The scenario generation process consists of two stages. In the first stage, a dynamic correlation network adaptively models temporal correlations by updating the covariance matrix, providing a more interpretable representation compared to black-box generative models. In the second stage, an implicit quantile network models the marginal quantile function in a nonparametric, continuous manner, enabling scenario generation through marginal inverse sampling. Experimental results show that the proposed dynamic correlation quantile network outperforms state-of-the-art methods in quantifying uncertainty and capturing dynamic temporal correlation. The source code is publicly available at: https://github.com/DREAMDXC/Dynamic-Correlation-Quantile-Network