Computer automated multi-paradigm modelling for analysis and design of traffic networks

In this article, Computer Automated Multi-Paradigm Modelling (CAMPaM) is presented as an enabler for domain-specific analysis and design of complex systems. Traffic, a new visual formalism tailored to the domain of vehicle traffic networks, is introduced. In the CAMPaM approach, the syntax of Traffic models is meta-modelled in an appropriate formalism such as Entity-Relationship Diagrams. From this description of abstract syntax, augmented with concrete (visual) syntax information, an interactive, visual modelling environment is automatically generated. The semantics of the Traffic formalism is subsequently modelled by mapping Traffic models onto Petri Net models. As the abstract syntax of models, irrespective of the formalism they are described in, is graph-like, graph rewriting can be used to transform models. Graph Grammar models thus allow for the specification of model transformations. The meta-modelling and transforma-tion of the Traffic formalism uses our CAMPaM tool AToM A Tool for Multi-formalism and Meta-Modelling. The advantages of creating a domain-specific formalism such as Traffic as opposed to using a generic formalism such as Petri Nets are presented. We also demonstrate how mapping Trafc models onto Petri Net models allows one to employ the vast array of Petri Net analysis techniques. In particular, a Coverability Graph is automatically generated and conservation analysis is automated by transforming this graph into an integer linear programming specification which is subsequently solved by the lp_solve code