The approach is called the surrogate model optimization (SMO), which relies on optimizing a surrogate model instead of the original model that requires significantly more computing time. The proposed SMO method combines the Latin hypercube sampling (LHS) method and a statistical approach called the ...