Design and efficient surrogate-based optimization of a novel Carnot battery energy storage system

As a long-duration physical energy storage technology centered on thermal energy storage, Carnot battery （CB） can effectively mitigate the grid integration impact of high-penetration renewable energy. However, current research predominantly focuses on the heat storage side of CB, leaving a noticeable gap in the design of cold storage-type CB systems and the efficient optimization of their thermodynamic operating parameters. This study first proposes a novel CB system coupled with a liquefied natural gas （LNG） regasification process. By integrating LNG cold energy as the cryogenic heat sink for the Rankine cycle, the power-to-power （P2P） efficiency of the system is significantly enhanced. Furthermore, an artificial neural network-based surrogate model optimization strategy is developed. This approach substantially reduces the computational cost and iteration time associated with complex rigorous model optimization, while retaining robust capabilities to search for the global optimum. The results indicate that the hyperparameter-optimized surrogate model can accurately predict system performance indicators and thermodynamic constraints. Specifically, the relative error between the predicted P2P efficiency and the results from the rigorous model is less than 1%, and the absolute error for the minimum heat exchange temperature difference constraints is within 1.5 ℃. When applied to particle swarm optimization, the computation time per generation is reduced from 1007.4 s to a mere 7.8 s. Moreover, the optimized system achieves a P2P efficiency of 76.13%, representing a 10.97% improvement over the initial design. This confirms that the proposed strategy significantly accelerates solving efficiency while ensuring effective process optimization.