Research And Application Of Intelligent Decision Reasoning Technology

The reasoning machine is a important department of intelligent decision system, it can solve some complex and material question by used field knowledge and specific method. And it also can make the Decision System have intelligence and solve question like expert. But the decision information have several characteristic include uncertain,instability,disaccord in the process of decision. And these characteristic bring difficult with decision and reasoning. So it is a key that how to reason in the intelligent decision system based uncertain information.The thesis is supported by the National Disaster Reduction Department's project of"Environment and Disaster Monitoring and Forecasting Application of Small Satellites for Disaster Reduction System"and the sub-project of'national science and technology support plan'project of"Research on Disaster Emergency Decision-making Support and Remote Consultation Technology", mainly carries on the research of the knowledge presentation method of the disaster reduction field, discussing the designing and realization of disaster reduction intelligent decision reasoning system. Firstly, based on the research of the representation method of multi-level knowledge unit, a new kind of knowledge unit is introduced, which makes the use range of knowledge unit cover with the knowledge of disaster reduction. And the knowledge is saved by utilizing XML concept. In the next place, discussing the qualitative reasoning method including Fuzzy Reasoning,Rough Set Reasoning to the uncertain condition. Then depicting the method of multi-reasoning machine cooperation based on the above several reasoning method. The method gain decision result from the qualitative,quantitative,uncertain,fuzzy reasoning method, then the main reasoning machine analyzed the result and output the solution. Next, the thesis discussing the method of decision information fusion, the information is result of qualitative reasoning and quantitative decision calculation. The dissertation puts forward several information fusion model, for example, zone-based fusion,time-based fusion,risk-based fusion and other fusion model. Finally, the designing and realization of disaster reduction intelligent decision system were researched, and depicting the function and user and the framework of the system. The thesis emphasize the realization method of Case-Based Reasoning Machine, Certain Factor Reasoning Machine and Fuzzy Reasoning Machine, and discussing the control method of main and secondary reasoning and the realization of decision information fusion. Finally, display the system interface, the flow and result of reasoning based on idiographic instance.