The Research Of Earthquake Disaster Management Based On Case-based Reasoning

Earthquake disaster,as an emergency,with its destructive potential and wide range of effects,attracts attention all over the world.Especially those countries,such as Japan,China,Indonesia,Chile,etc.,which are on the seismic belt,pay more efforts in studying earthquake emergency management.Generally speaking,China's earthquake disaster emergency management level is relatively behind the world's average,so our country attaches great importance to the study of improving the level of earthquake disaster emergency management.Case-based reasoning is an artificial intelligence technique.It imitates human thinking mode,which means to use the previoussuccessful solution to solve the present and similar problem.The design of case-based reasoning professional system is one of the hot spots.In the processof earthquake disaster emergency management,if similar previous earthquakes could found,its solution and effects will undoubtedly help tp make more scientific and rational decision.Therefore,it is of great practical significance to apply case-based reasoning technology to earthquake disaster emergency management.Based on this,this paper will describe the research and design process ofcase-based reasoning earthquake disaster emergency management system.Firstly,through the collection and separation of earthquake cases,it extractssome factors affecting the managementemergency management bystudying the measures taked in those cases and their reasons.After sifting and rectifying the factorsthat other experts and scholars have pointed out,and finally the system factors have been decided in this paper.Secondly according to the four steps of CBR(case retrieval,case reuse,case revising and case study,preservation),the proper case representation and retrieval methods have been provided: This paper combines the generative representation and frame representation methodto represent the seismic case;and it takes the nearest neighbor algorithm to retrieve and improvesthe efficiencythrough combining K-means clustering algorithm;then it primarily achieves to retrieve in the virtual database through the established expert system.The system has been able to implement basic functions: looking for old cases and comparing new cases.By looking for old cases,you can find examples of old cases that match certain conditions.By comparing new cases,we can buring up the the case that is similar to the target case from case library in order to providing support for the solution of the new case problem.