The Research On Agent-Based Model Transformation

The agent-oriented methodology provides a number of high-level abstractions based on organizational concepts to analyze and model multi-agent systems, in which agent plays some roles in the organization and cooperate with each other in order to achieve design objectives. It presents high-level abstraction to model multi-agent systems in a natural way and therefore is helpful to control the complexity of systems to be developed. Such abstractions can be used to support the analysis and modeling of the systems. Caste is the abstract description of agents, which have the same structural and behavioral characteristics. It can act as the basic modular and component to design and implement multi-agent systems. Therefore, it is necessary to develop method to bridge the gap between organization-based model and caste-based model.The thesis presents a method to transform the organization-based model to caste-based model. We borrow the idea of MDA and take organization-based model as platform-independent model and caste-based model as platform-dependent model. The algorithms of transformation are designed by analyzing the relationships between the organization-based model and caste-based model. A software tool is developed and a case is studied to validate our method.