Optimization of Stochastic Discrete Event Simulation Models Using "AFO" Heuristic

Abstract

The optimal conditions evaluation in complex stochastic systems modelled through Discrete Event Simulation is often extremely costly in computational terms. Especially when the number of variables involved is high, as in the case of manufacturing systems, the duration of each simulation run can last even several hours of calculation. It therefore becomes very important to use an optimal search method that allows the experimenter to reduce as much as possible the number of function evaluations analysed. With this goal, the Authors compared the performance of a new nature-inspired Heuristic called Attraction Force Optimization (AFO), with those of traditional algorithms, applying these different methodologies to a real industrial case. The authors believe that the obtained results could be of great interest to the scientific community and the AFO heuristic can become a valuable reference for discrete event simulation-based optimization problems.