Simulation-based constraint generation with applications to optimization of logistic system design

We present a general framework for a combined optimization/simulation approach where constraints to be satisfied are identified from results of simulation evaluation of the proposed system alternative, and then these constraints are added to the optimization model for re-optimization. The proposed cutting-plane-like procedure is iterative and terminates when an "optimal" solution of a mathematical program is obtained which passes all conditions of performance criteria set for simulation evaluation. A case of the real large-scale logistic system design is taken as an example, and the proposed approach is shown to work efficiently for the case, and looks promising for other problems, especially in the field of logistics and scheduling.