Exploiting Variance Behavior in Simulation-based Optimization

The methodological contribution proposed herein arises from considering the integration of stand-alone optimization techniques in discrete-event simulation, in order to model dynamic logistic processes, under realistic conditions of uncertainty. A simulation-based optimization approach is investigated to optimize overall system performance measures. A special focus is laid on evaluating alternative system solutions when they are either known a priori or revealed at run-time. To this end, a variance-guided statistical technique for the ranking and selection of candidate solutions has been devised and integrated into a solution generating algorithm on which the search process for the best solution may be centered. The findings returned from this work have been coupled with a queuing network model developed and applied in container stacking/retrieval operations via dts direct transfer system on the yard of a real maritime container terminal for pure transshipment.