Research On Multi-Agent Cooperative Framework Based On Pervasive Computing

Pervasive computing, combining of information space and physical space, is a kind of innovatory computing mode. In this combined space, the digital service can be pellucidly gotten anytime and anywhere. Agent has the character of autonomy and collaborate in distributed environment, and also has the ability to resolve knowledge and questions.With the development of distributed artificial intelligence and network technology, people bring forward more demands of resource sharing, distribute collaborate and operation wherever and whenever you want. It is the purpose of modern industry design to develop collaborative designing environment based on pervasive knowkedge and intelligence for simulating and helping human to deal with knowledge and information, so that designing process can achieve automatization.In order to reach the purpose and resolve the problem of modern collaborative framework, this paper presents a multi-agent collaborative framework based on pervasive computing, which resolves the problem that operation cann't be performed at anytime and anywhere, so that the pressure of multi-agents is alleviated and the safety of the system is incresed.Firstly, on the basis of pervasive computing technology and multi-agent technology, this paper enhanced the multi-agent architecture and collaborative framework based on pervasive environment. Then, a new layer of agent system is given, which is introduced to the multi-agent collaborative framework based on pervasive computing. What's more, the design theory and function of every agent are discribed in detail. To deal with the multi-channel information, there should be channel identification module, context generation module and multi-channel information integration module in the agent. In addition, Knowledge representation and storage based on design context are addressed in detail. This paper also discussed about the knowledge sharing approach among multi-agents in cooperative design system.The requirement of identity validation is presented to make pervasive computing resources safer. In other words, validation should not only work just in anytime and anywhere, but also require higher exactness. Then by leveraging context awareness techniques, the validation model, validation algorithm and validation principle are proposed.