A domain specific language for distributed modeling

As the digital transformation of industry continues, more and more data is being collected to gain insights into and further improve existing processes, known as prescriptive analytics. Among the enabling technologies for prescriptive analytics is simulation-based optimization. To accelerate the execution of simulations, the approach can be coupled with machine learning methods to create so-called surrogate models. However, this can lead to a loss of modeling accuracy if processes can only be inadequately mapped to such models. In this work, we present a new domain specific language, to model complex systems as a directed graph of smaller, communicating system components. With this language, surrogates may be developed more flexible, i. e., only for those parts, where it is meaningful. Further on, the execution of modeled components can be distributed to gain speedup. We provide an overview of the created language syntax, development process and support. We also show the applicability of the language in a case study: in terms of parsing speed, the language performs at the same level as comparable markup languages, while it outperforms them in terms of brevity, showing that it is more expressive. Finally, we outline additional features and the future application context of the language.