Interaction Structure of Multi-Agent System Based on Generalized Correlation Coefficient
基于广义相关系数的Agent行为决策模型

The study of interaction between agents is one of the most important areas in the research on multi-agent system. The process of decision making is such a key problem in the interaction that attracts much more endeavor. By now, the decision-making processes are limited to the specific environment or conditions, such as cooperation, negotia- tion, competition and so on. The framework of interaction based on generalized correlation coefficient unifies the all kind of interaction, from collaboration, selfishness to competition and antagonism. This paper is based on this frame- work. A model of agent based on generalized correlation coefficient is constructed. Furthermore, the process of deci- sion making based on this is analyzed. This process is more agile and more appropriate to the open and dynamic envi- ronment.