This paper makes analysis on theories of D-S evidential reasoning, rough sets and extenics sets, the three of which hold similarities in defining uncertainty sets. It’s feasible to introduce relevant theories and methods of extenics into information fusion and a method of extenics fusion (MEF) is presented as well. The method combines extenic correlation function with Dempster method and is considered to be a good solution in solving evidence collision and BPA function in information fusion method basing on D-S evidential reasoning. With the difference of transfer functions, the paper designs a method of extenics fusion. The simulation test results show that MEF is better than the traditional D-S evidential reasoning and is applicable to all kinds of evaluation problems.

This paper makes analysis on theories of D-S evidential reasoning, rough sets, and extenic sets; and indicates that there are some similarities among the three theories in defining uncertainty sets. Therefore, it’s feasible to introduce relevant theories and methods of extenics into information fusion and a method of extenics fusion (MEF) is presented as well. The method combines extenic correlation function with Dempster rule and is considered a good solution for problems of evidence collision and BPA function in information fusion method based on D-S evidential reasoning. The simulation test shows that MEF is better than the traditional D-S evidential reasoning and is applicable to assess all kinds of problems. Using the method of this paper to evaluate surface water in one area of Northern China, the results were consistent with the fact.