Dynamic Optimization Control of Virtual Power Plant with Seasonal Hydrogen Storage: An Energy Operation Method Based on Forecast Accuracy Assessment

Abstract

This study focuses on the integration of renewable energy sources such as photovoltaics and wind power into the low-carbon transition of power systems. The virtual power plant technology is key to enhancing distributed renewable energy efficiency, effectively integrating diverse distributed sources. To address the challenges of unknown uncertainties associated with the aggregation of dispersed renewable resources in VPPs, a novel energy operation method that incorporates seasonal hydrogen storage is proposed. In the method presented in this study, the L´evy α-stable distribution is utilized to construct the probabilistic model for the prediction errors of aggregated distributed energy, thereby creating dynamic scenarios across the long-term time scale. For the single-day optimization operation within dynamic scenarios, a dual-stage optimization strategy for hydrogen storage virtual power plants is proposed. In the day-ahead stage, a mixed Nash strategy weighting method is used to optimize both reliability and economic efficiency. In the real-time stage, hydrogen storage backup capabilities are employed to offset forecast errors, and information entropy is utilized to evaluate forecast data for dynamic optimization. The research results indicate that the method proposed in this paper provides essential flexibility backup for the virtual power plant over the long-term scale.

Keywords: Power system long-term planning, Virtual power plant, Seasonal hydrogen storage, The long term timescale, Dynamic scenario