Guangzhou Zeng

Learn More
This paper presents a construction approach for software agents. A software agent is regarded as a main frame plus some components, and it is constructed by selecting suitable main frame and software components, assembling and running through the control mechanism on the main frame. The multi-agent system is organised on the Client/Server model. The(More)
This paper proposes a domain-specific software architecture style. In the architecture style we define three software component types: procedure components, object components and agent components. We design a software bus or connector called Tribus to facilitate interactions among these components and specify one way quotation relations between different(More)
—Optimal workplace coalition generation is a key technique in migrating workflow. A trust model that describes cooperation relationships among workflow entities is presented. Different from traditional trust models, the trust model organizes confidence of a workflow engine in a workplace by two categories, capability confidence and recommendation(More)
In multi-agent system, the learning of actions selected by each agent results in poor cooperation, at the same time the conventional reinforcement learning requires a large computation cost because every agent must learn separately. A novel multi-agent cooperative learning model (MCLM) and a multi-agent cooperative learning algorithm (MCLA) are presented to(More)