When researchers generate a model for an experiment, they think of the model holistically, understanding which model pieces are important to answer their research question. Currently, the only way to generate such a model is to bring together numerous domain experts who each knows how to use his or her domain’s tool to generate that specific piece of the larger model. This approach is inefficient and resource intensive.
Arion is designed to alleviate these problems. It is a high-level modeling language and library that enables researchers to model their problem naturally. Arion converts the model into the languages of domain-specific simulators so that researchers don’t need to rely on domain experts to start their experiment.
The Java programming language is used to provide a suite of capabilities, but Arion is designed to make you feel like you’re defining a model, not performing programming.
Today, Arion supports GridLAB-D, ns3, and Framework for Network Co-Simulation (FNCS). Support for Opal-RT, EnergyPlus, VOLTTRON™, and powerNET is planned.
Arion creates the necessary configuration files so that researchers can go straight from modeling to receiving simulator data.
Arion is one of several projects under Pacific Northwest National Laboratory’s Control of Complex Systems Initiative (CCSI). CCSI is a five-year effort, bringing together theory, tools and testing of large-scale complex systems in a way that leads to effective deployment of more integrated, affordable and secure control systems. The initiative’s outcomes will benefit the power grid and buildings, with additional applications expected in transportation, national security, computing, manufacturing and other fields.