The AMoEBA Framework Enabling Formalized Execution of Physics-integrated Descriptive Architecture Languages
A core goal for an efficient and scalable simulation and analysis framework is the separation of the models, from the simulation execution, from the chosen methods for orchestration and/or analysis. In doing so, the analysis concepts, simulation architecture, and simulation execution can be abstracted and constructed in such a way to support inherent interoperability with diverse and best-of-breed models and tools assuming that the underlying context is well described and semantically grounded during data interchange. Consistent simulation context and semantic agreement of data objects produced and consumed during transaction is crucial in the design and construction of a model agnostic simulation infrastructure and orchestration framework. These concepts necessitate a set of abstract base classes which formally describe the overall physics phenomenology, operational domain concepts, systemic functions, and methods for process interaction. This formal descriptor captures the ontological relationships of objects and concepts and becomes the basis for integration and thus the very mechanism to assure that data relevance is preserved with respect to well-typed hierarchical acknowledgement and compositional/aggregational association in reverence to relational dependencies. Similarly, the common vernacular ensures consistency during object and data interchange and additionally subordinate meta-models guarantee unit transformation for efficient orchestrated execution.