Abstract

We present a simulation and optimization framework for stochastic resource-constrained project scheduling problems. The stochastic features are the job durations modeled by continuous random variables (without any time discretization) and the fluctuations are simulated by testing a sufficiently large number of realizations of an instance. The aim is to gain insights in the dependencies between the fluctuations of the input parameters and the objective function to enable a priori estimations. Such estimation methods developed by simulating small instances could be extrapolated to problems with a larger number of jobs or with more complicated features.