| The advent of multi-agent systems has brought opportunities for the development of complex software that will serve as the infrastructure for advanced distributed applications. During the past decade, there have been many agent architectures proposed for implementing agent-based systems, and also some efforts to formally specify agent behaviors. However, research on narrowing the gap between agent formal models and agent implementation is rare. In this thesis, we present a model-based approach to designing and implementing multi-agent software systems. Instead of using formal methods only for the purpose of specifying agent behavior, we bring formal methods into the design phase of the agent development life cycle. Our approach is based on the G-net formalism, which is a type of high-level Petri net defined to support modeling of a system as a set of independent and loosely-coupled modules.; We first introduce how to extend G-nets to support class modeling and inheritance modeling for concurrent object-oriented design. Then, by viewing an agent as an extension of an object with mental states, we derive an agent-oriented G-net model from our extended G-nets that support class modeling. The agent-oriented G-net model serves as a high-level design for intelligent agents in multi-agent systems. To illustrate our formal modeling technique for agent-oriented software, an example of an agent family in electronic commerce is provided. We show how an existing Petri net tool can be used to detect design errors, and how model checking techniques can support the verification of some key behavioral properties of our agent models. In addition, we adapt the agent-oriented G-net model to support basic mobility concepts, and present design models of intelligent mobile agents. Finally, based on the high-level design, we derive the agent architecture and the detailed design needed for agent implementation. To demonstrate the feasibility of our approach, we describe a toolkit called ADK (Agent Development Kit) that supports rapid development of application-specific agents for multi-agent systems. |
