Dynamic JChoc: A Distributed Constraints Reasoning Platform for Dynamically Changing Environments.

In Artificial Intelligence, a large number of problems (i.e. distributed resource management, distributed air traffic management, Distributed Sensor Network [1]) can be modeled and solved as Distributed Constraint Satisfaction Problems (DisCSPs). As many real world problems change continuously and incessantly over time, some methods have been developed (e.g. DynABT), for solving problems which exhibit this dynamic behavior. Meanwhile, there was no available framework that helped users to develope intelligent multi-agent systems based on Dynamic and Distributed Constraints Reasoning (DCR) techniques. In this paper, we propose a new platform, called JChoc, supporting the dynamic aspect for DisCSPs. JChoc is an easy to use platform, based on an elegant Multi-agent communication sub-platform (e.i JADE). It deals with agents with local complex problems and allows a realistic use of agents on a real distributed and dynamic framework. A real distributed problem is addressed to illustrate how the platform can be used to solve dynamically changing problems. However, the experimental results show the defectiveness of our platform.