Methodological assistance to a business analyst in studying the features of solving optimization problems in simulation modeling

Optimization techniques that are related: a) with ordinary mathematical problems based on linear and nonlinear models; b) with probabilistic systems using simulation models are fundamentally different from each other. In the article, the features of solving optimization problems in simulation modeling are studied on the model of a simplified version of the inventory management problem, in which the demand for products has a probabilistic nature. Statistical experiments on the simulation model confirmed the known positions that characterize the simulation results: a) are not stable due to the presence of a sample error; b) are represented not only by point, but also by interval estimates; c) are interpreted using statistical tests. Studies have shown that optimization using simulation modeling, in fact, is not a particularly reliable research method. The fact is that the accuracy of the simulation results is affected by both the duration of the simulation period and the number of runs of the simulation model during the statistical experiment.