Research On Rational Model Of Controllable Agent

VO can be considered as the ideal application environment for Agent technology. However the traditional Agent model based on modal logic still has some vital shortcomings for the application:credibility, computational complexity and conflict resolution.For the three questions,under the context of VO,based on policy-oriented mechanism,taking advantage of defeasible logic, to achieve the main purpose of controllability, we established a new rational model of controllable Agent in this thesis:First,individual Rational Model of Controllable Agent is proposed.Based on a two-tier policy-oriented mechanism, and taking advantage of defeasible logic, a new Policy-oriented autonomic,controllable Agent model is proposed,which Maintains a certain degree of autonomy while at the same time enhances the controllability in the organizational level.Second,organizational Collaboration Model of Controllable Agent is proposed. Using organization and policy metaphor, we build an organized and policy-oriented Agent collaborative model.Finally, validation example is given.Based on the formal theory of the model,we analyzed a specific example of VO.Taking one with another, the innovations of the model are as follows:First, using organization and policy metaphor, we build an organized and controllable Agent model.Through the policies,macro-guidance and external control are given to the Agents to enhance system dependability.Second,using extended Defeasible Logical framework,we establish a formal theory for the model which has linear time computational complexity.Finally, based on a two-tier policy-oriented mechanism, the Individual and collective conflict can be resoluted.