New Membership Transformation Method on Bridge Safety Fuzzy Evaluation

The key computation of bridge safety fuzzy comprehensive assessment using analytic hierarchy process is to realize membership transformation. However the existing membership transformation methods can't indicate "which parts in index membership are useful for object classification and which parts are useless". It results in that the redundant data of no use in object classification are used to calculate object membership. Use data mining based on entropy to find information in index membership about object classification. Then the relationship between object classification and index membership is clear. Distinguishable weight is defined to eliminate redundant data. Thus a new membership transformation method is proposed and used to evaluate bridge safety.