A Reliability Assessment Method Based on Environmental Stress and Multi-Source Information Fusion for Complex Product

A reliability assessment method based on environmental stress and multi-source information fusion is proposed for the problem of reliability assessment of complex products under small sample conditions. Firstly, considering the influence of temperature and humidity on complex products, the environmental stress acceleration factor is calculated based on the Hallberg-Peck model, and the working time in different environments is mapped to the reference environmental conditions. Secondly, multi-source information, such as historical test information and similar product information, are used to correct onsite use information. And according to the characteristics of different types of truncated data, the method of determining the corresponding distribution parameters is given. Then, the fitting factor is determined based on Bayes chi-square test and Bayes chi-square goodness of fit, and then the multi-source information is fused to carry out reliability assessment. Finally, the effectiveness of the proposed method is verified by an example analysis.