An Open Source Object- Oriented Architecture for discrete event simulation

Simulation is a valuable tool enabling organizations to investigate possible strategies for improving performances and reducing operating costs. In this work, an Object-Oriented Approach (OOA) to create simulation models and the relative open source software architecture are introduced. The goal of such an approach is to reduce model composition time by means of a direct system-model mapping. The resulting software has the following features: code reusability, modularity and extensibility. In particular, extensibility has a key-role to satisfy some common needs in the field of simulation model development. Moreover such an approach permits to perform fuzzy simulation run, consisting in modelling and simulating systems whose parameters are defined in uncertain form and no statistical distribution can be inferred because of the lack of data. In such a paper a description of the open source project is made, a comparison is presented between the presented methodology and other open source simulator projects in order to explain the main differences. Finally, an example in the manufacturing field is presented to show the potentiality of our approach.