Reliability Analysis Based On Performance Degradation Information

The improvement of manufacturing technology greatly enhances the reliability of products,which makes it difficult to obtain the actual failure data of products.Therefore,using product performance degradation data and other available information to analyze product reliability is an important issue in the field of reliability engineering in recent years.In this paper,reliability analysis and evaluation models were established for the performance degradation data in different environments.The main work is as follows:For the environmental performance degradation data of the products.firstly,the different conditions of the successively working environment of the product were studied.Hallberg-peck model was used to define the environmental factors,and the Wiener process degradation model of the product was established.The environmental factors are used to complete the transformation of the data in the working environment before and after,and the reliability analysis results of products are obtained by setting variable failure threshold.Secondly,considering that products are assigned to different environments,the evidence theory and Dempster's fusion rule are used to conduct data fusion of differences for each environment.The inverse Gaussian degradation process of the product is further established with the given failure threshold,and the reliability analysis results of the product been obtained.Numerical simulation and case analysis show the feasibility and effectiveness of the proposed method.A new method for modeling degradation data in multiple environments is proposed.A reliability analysis method based on data fusion is presented for products with performance degradation data and a small amount of life data.Firstly,a support vector regression model based on performance degradation data was established to obtain the pseudo-failure life data of products under a fixed failure threshold,and then the reliability analysis model was obtained.Finally,the model parameters were dynamically updated by combining the Bayesian theory and product failure life data.The comparison between simulation experiment and case analysis shows that the reliability evaluation of the proposed method is more accurate.