Building of an Architecture for the Fish Disease Diagnosis Expert System Based on Multi-agent

The paper presents the architecture of the MAS-based distributed fish disease diagnosis expert system on the basis of the analysis and summary on the characteristics of the existing distributed fish disease diagnosis expert system. That gives full play to the MAS-based distributed computing and to build flexibly. For heterogeneous knowledge sources of the distributed expert system, use the distributed computing agent to achieve the distributed diagnosis and management of fish diseases, and reduce the burden of the centralized processing. Meanwhile, it uses the modified contract net model, and designs a task allocation model of the collaborative diagnosis based on MAS for the distributed fish diseases, which describes the task allocation process concretely. Experiments proved that the system not only overcomes the limitations of single expert in the field, can also be cross-regional collaborative diagnosis. It improves the efficiency of the distribution of tasks and increases the accuracy of the results of fish disease diagnosis.