Multi-source data fusion and reliability analysis based on TPAC-KL model
Cong Wei, Fan Zhang, Peng Wang · 2024
With the increasing life span of aviation products, product reliability data is difficult to obtain and there are challenges in reliability analysis. In this regard, this paper proposes a multi-source data fusion method based on the TPAC-KL model. Firstly, under the two-parameter Weibull distribution model, the Maximum Likelihood Estimation of multi-source data is carried out to obtain the characteristics life of each type of data source; secondly, the TPAC Attribute Method is proposed to measure the credibility of the data sources, and the average similarity of the data sources is measured by the KL divergence; finally, the fusion of multi-source data is realized and the reliability analysis is carried out. The results show that the TPAC-KL model can effectively fuse the reliability information from different data sources, balance the differences between multi-source data, and improve the accuracy of the overall prediction of the target product reliability.