Researchers in China have proposed a scheduling framework for microgrids that integrates solar power with small modular nuclear reactors to improve short-term dispatch capability and long-term economic viability. The framework utilizes multi-objective distributionally robust optimization and real-time reinforcement learning to co-optimize photovoltaic and SMR generation. The system includes a generator, battery, electrolyzers for hydrogen production, and an energy management system to make decisions based on forecasts and real-time data. The proposed optimization framework reduces operational costs by 18.7% and carbon emission intensity by 37.1% compared to conventional fossil-dominated microgrids, while enhancing critical load supply reliability to above 98% across all uncertainty scenarios. Researchers have developed a new operational strategy for microgrids that combines demand response optimization with reinforcement learning to adapt to real-time environmental changes, allowing for flexibility in managing energy imbalances and reducing reliance on carbon-intensive backup generation. The coordination between short-term battery storage and long-term hydrogen storage enhances cost-effectiveness and reliability.
https://www.pv-magazine.com/2025/11/25/when-solar-meets-next-gen-nuclear/