A graph‐based computational framework for simulation and optimisation of coupled infrastructure networks

Abstract

The authors present a graph-based computational framework that facilitates the construction, instantiation, and analysis of large-scale optimisation and simulation applications of coupled infrastructure networks. The framework integrates the optimisation modelling package PLASMO and the simulation package DMNetwork (built around PETSc). These tools use a common graph-based modelling abstraction that enables them to achieve compatibility between interfaces and data structures and facilitates the modular creation and exchange of component models. The authors also describe how to embed these tools within complex computational workflows using SWIFT, which is a tool that facilitates parallel execution of multiple simulation runs and management of input and output data. Finally, the authors discuss how to use these capabilities to target coupled natural gas and electricity systems.