论文部分内容阅读
本文提出一种新的可靠性增长模型,用于分析开发试验过程中系统失效率。该模型只要求试验的每个阶段中实际存在但未知的失效分布是连续的,且只有一个未知参数θ和一个有关的单峰似然函数;对开发过程中失效率参数形式不作任何假设,只要求在设计变动期间,系统的可靠性不退化。在模型中,假设从一个试验阶段到下一个试验阶段的参数序列为:θ_1≥θ_2≥…≥θ_m。根据相对误差和均方差,说明新模型的性能很好。不考虑实际潜在的失效过程,总的来说该模型优于通常所用的AMSAA模型,且结果表明,在系统开发早期,AMSAA模型存在一个明显的偏差。
In this paper, a new reliability growth model is proposed to analyze the system failure rate during the development of the test. The model only requires that the actual existence but the unknown failure distribution in each stage of the test be continuous, and there is only one unknown parameter θ and a related unimodal likelihood function. There is no assumption about the failure rate parameter form in the development process Requirements during the design changes, the system reliability does not degrade. In the model, it is assumed that the parameter sequence from one experimental stage to the next one is: θ_1≥θ_2≥ ... ≥θ_m. According to the relative error and mean square error, the new model shows good performance. Regardless of the actual potential failure process, the model is generally superior to the commonly used AMSAA model and the results show that there is a significant bias in the AMSAA model early in system development.