A research and application of mechanism of multi-agent cooperative inference

The formal system of a multi-agent system based on modal logic and nonmonotonic logic are inadequate for complex objects in open and dynamic environments. The basic structure of the agent is analyzed in the paper. A cooperative inference system of a multi-agent system is constructed too. Agents cooperate to solve problems by different inference abilities and different knowledge bases of heterogeneous agents. The reliability is analyzed also. This method improves the inference ability of an agent and makes the system robust and reliable. When the multi-agent system designed in the paper is applied to analyze and adjust the components of a nutritional solution, the primary disadvantages of traditional methods are overcome