MULTI-SOURCE HETEROGENEOUS DATA FUSION MODEL FOR RELIABILITY EVALUATION BASED ON APPROXIMATE FAILURE POINTS

In order to solve the problem of multi-source heterogeneous data fusion and improve the accuracy of reliability evaluation, a multi-source heterogeneous data fusion method based on approximate failure point was proposed by using D-S theory and the least squares method.Firstly, the probability envelope curves were obtained through constructing probability assignment for a single source of data and performing weighted fusion.A distribution fitting model based on approximate failure points was also established.Secondly, the parameter estimation value was obtained by the least squares method, and the area metric was constructed to determine the true failure distribution function，and then the reliability assessment was completed.Finally, the feasibility and effectiveness of the proposed method were verified by examples, and the accuracy was higher than that of Bayes method.