Since aero-engine state monitoring data is fuzzy and random ,a method of improved information fusion is proposed based on the theory of Bayesian rough set and D-S evidence theory which can be used for aero-engine state assessment .Firstly,the test data is discretized and we extract the key parameters that effect aero-engine performance by using attribution reduction algorithm based on discernibility ma-trix.Secondly ,the Bayesian rough set and the support degree are redefined ,and the certainty gain function of the Bayesian rough set in the best decision table are calculated .According to the normalization method , the basic probability assignments are obtained .Finally,the evidences are fused utilizing the D-S combi-nation rule.This proposed method is applied to solve the problems of aero-engine state assessment ,The calculation results indicate the validity of practical application and the accuracy is enhanced in compari -son with other methods on the condition that information is insufficient .
NETL
NSTL
万方数据
维普
