Multi-objective optimization method for reactor shielding design based on SMS-EMOA

•First, the calculation method used two-dimensional SN code instead of Monte Carlo method. SN method has unique advantages in dealing with deep penetration problems.•It has fast calculation time and can complement Monte Carlo method;•Secondly, a multi-objective optimization method for reactor shielding design combining the SMS-EMOA algorithm and neural network was proposed, and the Pareto optimal frontier was obtained;•Finally, the use of hypervolume indicator was added to evaluate the quality of optimized solutions.