Research On Financial Decision Support System Based On Multi-Agent

With the globalization process of market and the rapid development of information technology, the market competition becomes more and more fierce for the enterprises and the operating risk of enterprises keeps growing. As the necessary tool for the advanced financial managers, the financial decision support system plays an important role in helping the decision maker to make correct judgments in order to ensure the competitive predominance of the enterprises. But the traditional financial decision support system can't satisfy the current requirement because of the distributed financial decision data and the customized function needs of decision makers. Therefore, we use the multi-agent technology in the field of artificial intelligence to design a more adaptable and intelligent financial decision support system based on the current requirements.This paper starts with the research of financial decision support system in domestic and overseas, and analyzes the deficiencies of existed system, and then proposes the necessity of introduction of multi-agent technology. Then the paper introduces the relative supporting theories and technologies for the following research. Secondly, this paper sets up the framework of multi-agent based financial decision support system on the base of functional requirement analysis. Then the paper describes the detail designs of the internal resource base and the function agents of the system, including the designs of data warehouse, model base, knowledge base, internal structure and the coordination mechanism of the agents. In order to coordinate the operation of the internal agents in the system, KQML is selected as the communication language among agents, and improved contract net model is selected as coordinating mechanism among agents.At last, this paper take the Faw Foundry which is one of the subsidiaries of China Faw group as the empirical researching subject of multi-agent based financial decision support system to verify the feasibility of the scheme in the paper and the practical value of the system proposed in this paper.