Dynamic Cooperative Multi-Agent Model of Complex System Based-on Sequential Actions' Logic
Jiang Hua wei · Chinese Journal of Computers · 2013
Dynamic complex problem solving has been an important issue of the theory and application of artificial intelligence and complex adaptive systems,where multi-agent cooperation is one of the hot and key points.And how to divide the problem and task needs more work.Benefiting the idea of histology,we Abstract self-determining running units of self-adaptive systems into agents,make complex adaptive system as a multi-agent organization.From the perspective of time and status,we describe system behaviors and propose multi-agent dynamic adaptive mechanism which is based on temporal activity logics.The mechanism contributes to complete dynamic cooperation tasks in multi-agent systems and built the task structure model.Also the semantic rules,authorizing rules and action rules for cooperative inference are built.A detailed description of the task to achieve dynamic collaboration is listed at six aspects: dynamic allocation of tasks,collaboration produce,collaborative group generation,common planning,collaborative groups,action and evaluation of results.To verify the feasibility and effectiveness of our approaches,we also do simulation on MAGE and other platforms.