Towards Automation of Simulation Studies

Abstract

Simulation is applied in various domains to investigate different variants, to study effects, or to optimize models with respect to some performance measurement. Simulation models can exhibit a high level of complexity and thus, consist of many components and potentially many adjustable input parameters as well as a large number of output measurements. In this habilitation thesis, different methods in the fields of automation, data mining, and optimization in the context of simulation are developed. The underlying motivation is to increase the degree of automation and to increase the efficiency when performing simulation studies. The proposed methods are evaluated using three different simulation systems: Manufacturing simulation, traffic simulation, and gas dispersion simulation.