System Reliability Method And Research Of Small Sample

At present, the reliability method is a better way to solve the uncertainties, but the amount of data in the actual project is very limited. And it is sometimes difficult to satisfy the reliability method on data quantity requirement. Therefore, in view of the limitation of reliability method in small sample application, on the basis of former, this article puts forward some improvement methods, and it is applied to reliability assessment of a certain type of hang tail. This article makes some preliminary discussion in the following several aspects, specific as follows1)Mathematical model of reliability assessment: take interval method and the probability method to calculate reliability smaller value for reliability analysis results, thus amended the stress - strength interference model of non-probability set. Reliability analysis results of the modified model is lower than the reliability of probabilistic model, and reliability assessment results is relatively conservative. Then, the amended model have been applied in performance degradation products of reliability assessment, and discuss how to determine variable interval according to the existing data. The result shows that: when appling the most value method in initial stage will highly estimate the product reliability. Therefore, in the early this method should be avoided. When application of interval statistics and interval gray number method used in the early and medium-term of reliability obtains more conservative, in the latter part of the reliability obtains the same with probability model.2) the function relations between the safety factor and the probability reliability measure: the article discusses the function relations between the safety factor and the probability reliability measure, demonstrating the intrinsic link and differences between the two. The results show that probability reliability assessment method with the probabilistic safety coefficient between the theory is mutual equivalent, there exist differences. Difference between them is mainly manifested in the design thought, measure and performance forms. Generally speaking, both focus and solving process are different, but both in solution is consistent with the conclusions.3) hang tail structure strength small sample reliability assessment: The probability and non-probabilistic method is applied to reliability analysis models into the tail intensity, and comparing the reliability assessment results between probability and non-probabilistic, discussing the method of choosing variable interval. Finally, using pseudo-experimental data on a model of the tail intensity of non-probabilistic reliability simulation. The results show that the modified model can do well on product reliability evaluation. Using grey variable interval range is superior to the statistics range, and the gray method can better eliminate abnormal data. Using pseudo-experimental data on a model of the tail intensity of non-probabilistic reliability simulation, it gives the small sample general steps and methods in the actual project reliability assessment.