基于无失效数据的轴承可靠性研究

发布时间：2018-01-13 23:11

本文关键词：基于无失效数据的轴承可靠性研究　出处：《东北大学》2014年硕士论文　论文类型：学位论文

【摘要】：随着轴承生产制造水平的发展,轴承寿命有了很大的提高。在轴承可靠性试验中,少失效、单失效乃至无失效试验结果经常出现。在这种情况下,由于失效信息的缺乏,传统的评价方法很难做出准确的可靠性评估,因此研究新的评估方法显得十分重要。长期积累下来的轴承寿命数据和相关理论可知轴承寿命是服从威布尔分布模型的,本文以此为前提,应用置信限法,贝叶斯方法,参数估计法等,对无失效数据情形下轴承的可靠性分析方法进行了深入的理论和仿真研究。首先,阐述了可靠性评价的目的及意义及轴承可靠性分析理论基础,并介绍课题来源,讨论轴承寿命分布,即威布尔分布相关参数和可靠性指标,介绍轴承的分组和单组定时截尾无失效数据类型,分析置信限法的原理及适用范围。其次,论述用于可靠性评估的配分布曲线法理论和步骤,该方法应用的前提是先估计失效概率,本文提出四种估计失效概率的方法,通过各方法处理结果的比较,讨论四种方法的估计精度。再次,选用拟似然函数思想、修正的极大似然方法以及相关系数优化法分别处理轴承寿命服从两参数及三参数威布尔分布时的无失效数据,研究无失效数据问题的参数估计方法。在理论研究的同时,结合仿真算例进行验证。再后,结合计算机仿真模拟试验,介绍了抽样思想方法和大量模拟数据统计规律法的原理和步骤,结合该方法的适用范围,即单组无失效数据情形下,给出了评估轴承可靠性的新方法。依据轴承研究所试验数据的实例研究,讨论了三参数威布尔分布在轴承寿命分析中的应用。指出引进最小安全寿命含义可以更科学、准确地评估轴承寿命等可靠性指标,论证了轴承三参数威布尔寿命分布假设要优于两参数威布尔寿命分布假设。最后,总结全文研究的内容及结果,并提出了未来的研究方向与不足之处。
[Abstract]:With the development of bearing manufacturing level, bearing life has been greatly improved. Less bearing failure at the reliability test, and no single failure, failure test results often appear. In this case, due to the lack of failure information, the traditional evaluation method is very difficult to make accurate reliability evaluation, so it is very important to study the new method of assessment.
Bearing life long accumulated data and related theories of the bearing life is to obey the Weibull distribution model, in this paper, the application of confidence limit method, Bayesian method, parameter estimation method, the reliability of the situation of zero failure data bearing analysis method were used to carry out in-depth theoretical and simulation studies.
First of all, expounds the purpose and significance of the reliability evaluation and bearing analysis theory, and introduces the topic source, discuss the bearing life distribution, the parameters of Weibull distribution and reliability index, grouping is introduced and single bearing group censored failure data types, analysis of the confidence limit method principle and applicable scope.
Secondly, the theory and procedure of the distribution distribution curve method for reliability evaluation are discussed. The premise of this method is to estimate the failure probability first. In this paper, four methods to estimate the failure probability are proposed, and the accuracy of the four methods is discussed by comparing the results of each method.
Again, the quasi likelihood function, maximum likelihood method and correlation coefficient method with zero failure data bearing life obey two parameter and three parameter Weibull distribution, respectively, of parameter failure data estimation method. In the theoretical research at the same time, combined with the simulation example was verified.
Then, combining the computer simulation experiment, this paper introduces the principle and procedure of sampling method and simulation data statistics method, combined with the scope of application of the method, namely the single situation of zero failure data, a new method to assess the reliability of bearings is given. On the basis of test case study data bearing Research Institute, discusses the application of the three parameter Weibull distribution in the bearing life analysis. It points out that the introduction of the minimum safe life meaning can be more scientific, evaluation of bearing life reliability index accurately, the argument of Zhou Chengsan parameters Weibull distribution is assumed to be better than the assumption of two parameters Weibull distribution.
Finally, the contents and results of the full text research are summarized, and the future research direction and inadequacies are put forward.

【学位授予单位】：东北大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TH133.3;TB114.3

【参考文献】