Stochastic Generation of Discrete-Event Simulation Models

We present a stochastic model generator for creating discrete event simulation models. The size, structure, layout and behavior of the generated model can be parameterized. By using stochastic techniques in all sections of the generation process a huge number of different models can be created. These models are to be used to test and validate new methods in automatic model abstraction and simulation visualization. We illustrate the functionality of the generator by describing detailed algorithms, which can be used to generate models for common discrete event simulation tools. In the end of the paper we present results of an implementation in the simulation tool d(3)FACT insight which demonstrates the applicability of the generator.