Advanced Modeling of Complex Behaviour in Concurrent Systems

In this paper we motivate the use of a declarative programming style and the use of massive parallel computing systems to model control strategies in concurrent systems. The simpliied formulation of system components and their interaction permits to generate a rapid prototype of the system. Based on this prototype, a rst simulation of the modeled strategies and the system is done on a massive parallel computing system. Using parallel hardware the massive parallelism of real systems can be naturally matched to an execution model. To allow a sophisticated modeling and rapid proto-typing we use strategies based on fuzzy logic and present a prototype of a fuzzy library for the concurrent programming language PCN 4]. We demonstrate the modeling strengths of our approach by describing aspects of a 'real world problem', a color sorting assembly buuer used in car production.