Functional Hybrid Modeling from an Object-Oriented Perspective
Henrik Nilsson, John R. Peterson, Paul Hudak · 2007
The modeling and simulation of physical systems is of key importance in many areas of science and engineering, and thus can benefit from high-quality software tools. In previous research we have demonstrated how functional programming can form the basis of an expressive language for causal hybrid modeling and simulation. There is a growing realization, however, that a move toward non-causal modeling is necessary for coping with the ever increasing size and complexity of modeling problems. Our goal is to combine the strengths of functional programming and non-causal modeling to create a powerful, strongly typed fully declarative modeling language that provides modeling and simulation capabilities beyond the current state of the art: in particular, support for highly structurally dynamic systems. Additionally, we think our approach could serve as a semantical framework for studying modeling and simulation languages supporting structural dynamism, and maybe even as a core language in systems where the surface syntax is more conventional. Although our work is still in its very early stages, we believe that this paper clearly articulates the need for improved modeling languages and shows how functional programming techniques can play a pivotal role in meeting this need.