Novel method for fuzzy comprehensive evaluation based on set pair analysis

In the application of fuzzy comprehensive evaluation, it is quite difficult to build membership function. In order to avoid determining membership function in fuzzy comprehensive evaluation, a novel method, called fuzzy comprehensive evaluation method based on set pair analysis (SPA), is presented. In set pair analysis, an uncertain system is described by only three states; identity, discrepancy and contrariness. In our method, we use connection degree matrix of SPA instead of membership degree matrix in fuzzy comprehensive evaluation. We first build the set pair connection degree matrix of the main indexes of the evaluated system; then the connection degree matrix is obtained; finally, fuzzy comprehensive evaluation is carried out based on the connection degree matrix. An example is used for analysis with our method; the results show that our method is much more simple than fuzzy comprehensive evaluation, and that it is an effective method for evaluation and comparison of several similar uncertain systems or different design schemes of a system.