Accelerated Storage Reliability Model And Statistical Analysis Of Electronic Products

Storage reliability is a complex systems engineering. It is very important in economy, also in military affairs. If the storage period is short, it needs to replacing the correlative components, of course, which will bring the waste of manpower, material resources and money. Otherwise, if the storage period is longer than its intrinsic, which will affect the use of effect of products or bring fatal accidents. Thus many people want to know the knowledge of storage reliability.According to demands of international position and electronic equipment systems, electronic equipment must adapt to traits of long-storage and useable at any moment, so it presses for electronic equipment satisfying the least storage-life. Especially for the electronic equipment systems of long-storage and one-off use, which storage-life and storage reliability are important technical index, because these connect security-storage and effective use. Such studying storage reliability of electronic products is very important.To begin with, this paper applies accelerated life testing to storage process of electronic products, using TTT-plot method to dope out the distribution of accelerated invalid data. Furthermore we can get electronic products' reliability and dope out their life. Final we can obtain reliability and reliable life of normal condition's products. In the course of this, I consider the reliability model that associating with mission time.Paper includes preface and nine chapters, details are following:Preface , the background of the research has been introduced , bringing out technical way andmain methods of research.Chapter 1, summarizing of basic concepts and theories. This chapter introduces the background ofresearch and classes of accelerated storage testing, furthermore it also introduces acceleratedmodels and concepts of storage reliability.Chapter 2, brings out accelerated storage reliability models with initialized invalidation, then thischapter explains initialized invalidation,accelerated storage testing and meaning of acceleratedstorage reliability.Chapter 3, making use of TTT-plot method to dope out the distribution of accelerated invaliddata. The process is following: to begin with, doping out the distribution of invalid data with samestress; then, converting the invalid data of different stress to equivalent invalid data of samestress, doping out the distribution of all invalid data. Some examples (non-censored sample andcensored sample) concerned.Chapter 4, gives six accelerated storage reliability models in the condition of initialized invalid,besides, parameters' estimate given.Chapter 5, studying accelerated storage reliability models with masked invalid data; then thischapter making use of EM algorithms to analyzes the II -censored invalid data in acceleratedstorage testing. Finally parameters' estimate of NHPP given Chapter 6, introducing accelerated storage reliability models which combining storage anduse. Then this chapter gives statistical analysis for the "first-storage, second-use" reliabilitymodels. Finally we give a example to dope out reliability and life for the "first-use,second-storage" model.Chapter 7, we transform accelerated storage reliability into normal storage reliability making useof life-stress model. Then we give particular statistical analysis for the three models: temperatureis accelerated stress, humidity is accelerated stress and T-H is accelerated stress. Reliability andlife of normal condition given in the end of this chapter.Chapter 8, summarizing technical way to improve storage reliability of electronic products.Chapter 9, summarizing this paper.In the study, the research of electronic products storage reliability is extended, and new techniquehas gone on new exploration. The possible innovations are following: 1. Applying accelerated lifetesting to storage reliability; 2 . Making use of TTT-plot method to dope out the distribution ofaccelerated invalid data; 3 . studying accelerated storage reliability models with masked invaliddata making use of EM algorithms.The contents and fields remained deep discussion include: 1. Whether testing processes consume electron products' life or not, because testing is equivalent to use; 2 . How to get electronic products' storage reliability and life without knowing invalid data.