Estimation Of Environment Factor In System Reliability And Software Realization

The reliability analysis and life prediction of highly reliable and long-life products are hot issues, how to get enough lifetime data under working environment has become an important research direction. After a long period of research,it’s find out that more data could been obtained under high stress by accelerated life testing. Then environment factor are raised to solve the problem of data’s conversion between high stress and normal working environment. It can greatly increase the data under working environment by environment factor, which discounting data of high stress to normal stress according to some rules. Thus the analysis of products’ reliability or prediction of product’s lifespan based on these data can be achieved. This paper mainly studies the method to estimate the value of environment factor based on degradation data, the problem of environment factor’s prediction based on degration data under different tempreture, and the method to predict lifespan of products based on environment factor. The main research results are as follows:1.Estimating environment factor based on degradation testing. In this paper, advantages for degradation life testing are analyzed compared with traditional one, and detailed information of failure-censored testing and time-censored testing is introduced. The characteristics of degradation data has been studied and illustrate the rationality of using them to to estimate environment factor. This paper combines method of BLUE(best linear unbiased estimation) with the distribution properties of degradation data to estimate the value of environment factor. An example is given to illustrate the estimation process in the end.2.Estimation of environment factor based on degradation data and accelerated testing. Under high stress products usually degenerate more quickly, based on this character this paper combines failure-censored degradation testing with accelerated testing to estimate environment factor. In this paper, only temperature is considered, so method of building environment estimation model and estimating its parameter is raised based on Arrhenius model and degradation data. To prove its accuracy, a new method is raised by comparing the degradation path which has been discounted by environment factor with the one without being discounted. Finally, it finds out that this method is accurate.3.Life predicting based on environment factor. This paper raised a new method to predict lifespan of products based on degradation data of high stress through accelerated testing, which uses model of environment factor to estimate environment factor under working temperature, then acquiring degradation data under working stress and correcting these data by the real degradation data of products under working condition. After constant integration, degradation data acquired by estimation could be more accurate. When there are a group of products working under normal temperature, this paper gives the way to estimate their average life, and when there is only one product, this paper raise a way to estimate its failure life. The final example illustrates the detailed process, whose results prove the feasibility of this method.4.The software system with environment estimation and life prediction is designed and implemented. By using Matlab, value of environment factor can be analyzed and calculated by this system, and life of products can be predicted. This system can also acquire products’ degradation path, judge the distribution of degradation data, achieve life information of products and validate the precision of estimation results.