Multi-objective optimization of carbon emissions and economic benefits in green hydrogen-coupled dimethyl oxalate hydrogenation process

Current research on green hydrogen-coupled hydrogenation systems primarily focuses on process design optimization, while neglecting long-term online operation optimization studies. Given the significant carbon reduction potential and economic viability of green hydrogen-coupled dimethyl oxalate (DMO) hydrogenation process to synthesize ethylene glycol, this study proposes a multi-objective optimization framework considering green–gray hydrogen ratio fluctuations. First, a new multi-objective problem considering carbon emissions and economic benefits under hydrogen ratio fluctuations are formulated. Subsequently, a first-principles model (FPM) is developed, with critical operation parameters and ranges identified through sensitivity analysis. And a high-fidelity offline surrogate model of FPM is established using Latin hypercube sampling and Gaussian process (GP) to generate the initial population of online surrogate model. Finally, we propose a surrogate-assisted optimization algorithm (APB-NSGA-II) to solve the above constructed multi-objective optimization problem, integrating adaptive-parameter GP, Pareto-based bi-indicator infill sampling into fast non-dominated sorting genetic algorithm (NSGA-II). In a three-month green hydrogen-coupled DMO hydrogenation simulation, compared with fixed-parameter strategy, APB-NSGA-II increases economic benefits by 1,383,773 yuan and reduces total carbon emissions by 2,027.7 tons; compared with NSGA-II, APB-NSGA-II increases economic benefits by 549,900 yuan and reduces total carbon emissions by 835.2 tons. This framework not only addresses operation optimization challenges in green hydrogen-coupled DMO hydrogenation process under hydrogen ratio fluctuations, but also provides methodological guidance for other hydrogenation processes integrating renewable hydrogen.