Research On Typhoon Case-based Reasoning Model Based On GIS

Case-based reasoning technology has been developed rapidly in recent years.It is an important knowledge-based solution and learning method of artificial intelligence.It solves the problem by reusing or correcting past solution of the similar problems.It gets the source case form the history case library by using the suggestive information of the target case,and uses the solution of source case to guide solving the current case.With frequent disasters in recent years,the case-based reasoning have been applied to the relief decisions more and more,and yielded the certain result.But they rarely think about the space information of disasters in case-based reasoning process and few people apply GIS to the case-based reasoning.So this paper mainly researches the case-based reasoning model of typhoon disaster combining with spatial analysis function of GIS.Firstly,through the analysis of the study,domestic and foreign research present situation of case-based reasoning,it indicates the importance and practical applicability of the GIS combining with case-based reasoning to solve problems,and introduces the related technology,some algorithms of case-based reasoning and the advantages and disadvantages of these algorithms.Secondly,according to the characteristics of spatial and attribute data of typhoon,the object-oriented attribute representation method has been applied to represent the case.And we establish the case library of history typhoon.In case retrieval,we use the spatial similarity algorithm which based on the GIS spatial analysis function and attribution similarity algorithm to search the similar history typhoon.So it could avoid comparing with all the cases,and improve the retrieval efficiency.Meanwhile we make the algorithm more suitable for the typhoon track and improve the accuracy of similarity by using the line buffer analysis.In order to make the solution of similar case more suitable for the current typhoon case,the genetic algorithm is used in case revising,and the relief supplies information of current typhoon is deduced.Finally,we utilize C#language to develop the case-based reasoning model and verify the validity of the model by the experiment,apply the theory to practice.So it provides auxiliary support services for natural hazard reduction decision-marking.