Research On Intelligent Matching Method Of Emergency Decision-making Of Freeway Incidents

Freeway traffic accidents occur frequently.Rescue workers making timely and effective rescue manners plays an important role in the best to save life and property security.In this paper,intelligent matching methodson emergent rescue decisionson freeway incidents are studied.Based on case-based reasoning technology,two kinds of case retrieval algorithm are proposed.Firstly,every part of the CBR system processof emergency decisions on freeway incidents is made the detailed description in this paper.In order to further study methods of case retrieval,a kind of case representation based on XML is proposed.Cases of the system are represented in this way.Then methods of case retrieval are researched from the two aspects:one is based on the similarity of cases,which excavates the differences between the attributes fully and covers the shortages of incomplete information effectively,and combines multiple attribute similarity with structural similarity to calculate the case similarity between incidents and history cases by the optimized nearest neighbor algorithm,and join case confident degreein the reuse link of case retrieval to get the best matching case.Another is based on the structural similarity and classification theory of naive Bayes,whichdivides thecase base in the first place and judge the categories of incidents by thestructural similarity,and then establish a classification model based on naive Bayes to get the best classic case with maximum posteriori probability.This method solves the problem of lack of attributes.In the end,the paper introduces the software design and process of implementation of the system which can select automatically the appropriate matching algorithm to output the best matching case.By simulation examples to verify expectations of the theoretical analysis in this paper and to show that this intelligent matching methods are effective and feasible.