Driven by global climate change and carbon reduction targets, nuclear energy has gained increasing prominence as a clean baseload power source. Enhancing the energy efficiency of key equipment in nuclear power plants is essential for achieving a low-carbon transition. This study addresses the trade-off between separation efficiency and pressure drop under multi-parameter coupling in hooked corrugated plate separators by proposing a multi-objective optimization strategy that integrates automated numerical simulation with data-driven optimization. An automated CFD framework was developed to efficiently generate a comprehensive dataset covering inlet velocity, droplet diameter, plate spacing, and hook length. A multilayer perceptron (MLP) surrogate model was then constructed, achieving high predictive accuracy with coefficients of determination (R2) of 0.95 for separation efficiency and 0.91 for pressure drop. Using the trained surrogate model, the NSGA-II algorithm was employed for multi-objective optimization, and the TOPSIS method was applied to identify the optimal compromise solutions. The results show that for representative droplet diameters of 5, 10, and 15 μm, the optimized structures improve separation efficiency by 25.71–29.14%. The integrated automated CFD–surrogate model–multi-objective optimization framework established in this study provides an efficient and generalizable approach for the design of gas–liquid separation equipment, contributing to energy consumption reduction in nuclear and process industries and supporting the realization of global carbon neutrality goals.
Gui et al. (Thu,) studied this question.