序进应力加速寿命试验加载方式及其统计分析方法的研究

发布时间：2019-02-13 11:13

【摘要】：加速寿命试验被广泛应用于研究产品的可靠性，序进应力加速寿命试验（简称序加试验）由于试验应力持续提高，能更快地加速失效，试验效率高，已成为加速寿命试验技术的一个重要发展方向。目前国内外对序加试验的研究主要集中在试验的优化设计与统计建模上，尚未对序加试验的加载方式做过系统的研究，以期获得更好的试验效果；也没有就如何用步加试验逼近序加试验开展研究，一方面可以避免序加试验对设备要求较高的缺陷，另一方面又可以达到序加试验效果；更没有就如何利用序进应力加速试验获得比最优恒定应力加速试验更高的估计精度展开研究。因此，为了解决序加试验应如何加载以及试验数据如何分析的问题，本文以航天电连接器为工程背景，以寿命服从威布尔分布、满足逆幂律或阿伦尼斯方程的产品的为例，系统地研究了序加试验的加载方式与统计分析方法，以获得易实施、效率高且估计精度高的加载方式，并且给出了更准确的参数估计方法，以期为电连接器等高可靠、长寿命产品的序加试验的开展提供技术指导。本文各章主要内容为：第一章，阐述了本文的研究背景与意义，然后从试验统计分析、优化设计与工程应用的角度简要地介绍了目前加速寿命试验的主要研究现状。而后又介绍了当前恒定、步进与序进应力加速寿命试验的主要研究内容与不足之处。最后介绍了本文的主要研究内容。第二章，首先给出了本文的基本模型假设，为了使统计模型容易建立，采用了转化应力线性增加的序加试验，并给出了常用的三种不同的加载方式，然后推导出了这三种应力下产品的寿命分布函数，最后给出了相应的模型参数的估计方法。第三章，以设计应力下对数分位数寿命的极大似然估计量的渐近方差最小为目标，对本文中的三种序加试验进行了优化设计，并研究了设计参数的错定对优化效果的影响。对三种加载方案从估计精度、稳健性与数据分析的复杂程度等方面进行了比较，结果显示三种方案各有所长，但综合考虑，方案二即转化应力线性增加且有转折的加载方式最好，其估计精度高、加速效果好且易于实施与控制。第四章，基于蒙特卡洛法，给出了序加试验的模拟方法，针对极大似然估计法在小样本时误差较大的问题，利用试验模拟的方法研究了小样本时极大似然估计法的误差，给出了修正方法。第五章，从加速效果与参数估计精度的角度对步加试验与序加试验进行了对比，得出当步加试验的载荷步数达到十数步时其与序加试验几乎等效，从而有效地解决了序进应力加载困难的问题。第六章，，研究了恒加与序加之间的关系，结果显示最优恒加试验可视为序加试验应力上升速率为0的特殊情形；鉴于最优恒加试验由两组应力水平不同的试验组成，其可被视为包含两组不同的序进应力的序加试验的特殊情形，对这种序加试验进行优化设计，得到了估计精度比最优恒加试验更高的恒加与序加混合的加速寿命试验，其中序加部分可用数步的步加代替。第七章，总结了本文的主要研究内容，提出了进一步的研究方向。
[Abstract]:The accelerated life test is widely used in the research of the reliability of the product, and the accelerated life test of the sequence in the order (the sequence addition test) can accelerate the failure more quickly due to the continuous improvement of the test stress, and the test efficiency is high. It has become an important development direction of the accelerated life test technology. At present, the research on the sequential addition test at home and abroad is mainly focused on the optimization design and statistical modeling of the test, and the loading mode of the sequence plus test has not been studied systematically, with a view to obtaining a better test effect, and no research is carried out on how to use the step-plus-test approach and the test. On the one hand, it is possible to avoid the defect that the sequence-plus-test has higher requirements on the equipment, and on the other hand, it can reach the order-plus-test effect, and it is more difficult to study the estimation precision which is higher than the optimal constant-stress acceleration test by using the sequence-in-stress acceleration test. Therefore, in order to solve the problem of how to load the order-plus-test and how to analyze the test data, this paper takes the space-electric connector as the engineering background, and the life is subject to the Weibull distribution, and the product of the inverse power law or the Arrhenius equation is satisfied. The loading method and the statistical analysis method of the sequence-plus-test are systematically studied, so as to obtain the loading mode which is easy to implement, high in efficiency and high in estimation precision, and a more accurate parameter estimation method is given, in order to provide the electric connector and the like with high reliability, The sequence and test of long-life products shall be provided with technical guidance. Main contents of the chapters of this paper In the first chapter, the research background and significance of this paper are described in the first chapter, and then the main research of the current accelerated life test is introduced from the angle of the test statistic analysis, the optimization design and the engineering application. The main contents and contents of the current constant, step and sequential stress-accelerated life test are also introduced. At the end of the paper, the main research of this paper is given. In the second chapter, the basic model assumption of this paper is given. In order to make the statistical model easy to establish, the order and test of the linear increase of the transformation stress are used, and three different loading methods are given, and then the life of the products under these three stresses is derived. The function of life distribution is given, and the corresponding model parameters are given. In the third chapter, the optimal design of the three-order addition test in this paper is carried out with the minimum of the asymptotic variance of the maximal quasi-linear estimator of the life of the log sub-digit in the design stress, and the wrong determination of the design parameters is also studied. The influence of three kinds of loading schemes from the estimation precision, the robustness and the complexity of the data analysis is compared, and the results show that the three schemes are different, but the scheme 2 is that the transformation stress is linearly increased and the loading mode with the turning point is the best. The method has the advantages of high estimation precision and good acceleration effect, It is easy to implement and control. In the fourth chapter, based on the Monte-Carlo method, the simulation method of the sequence-plus-test is given. In the light of the problem that the error is large in the case of a small sample by the method of the maximum likelihood estimation, the error of the method is studied by means of the method of the test simulation. In the fifth chapter, the step-plus-test and the sequence-plus-test are compared with the angle of the acceleration effect and the parameter estimation precision, and it is concluded that when the load step number of the step-plus-step test reaches the ten-step, it is almost equivalent to the sequence-plus-test, thus effectively solving the problem. In chapter 6, the relationship between constant and sequential addition is studied. The results show that the optimal constant-load test can be regarded as a special case with the increase rate of the stress rising rate of the test and the test. In view of the optimal constant-load test, The test composition of two groups of stress levels can be considered as a special case of the sequence-plus-test of two groups of different sequence-in-stress. In chapter 7, the main research of this paper is summarized.
【学位授予单位】：浙江理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TB114.3

【参考文献】