计算机集成制造系统 ›› 2015, Vol. 21 ›› Issue (第8期): 2124-2131.DOI: 10.13196/j.cims.2015.08.018

• 产品创新开发技术 • 上一篇    下一篇

基于质量特性演化的工艺可靠性评估

匡芬1,戴伟1+,王健2,赵宇1   

  1. 1.北京航空航天大学可靠性与系统工程学院
    2.北京无线电测量研究所
  • 出版日期:2015-08-31 发布日期:2015-08-31

Process reliability evaluation based on quality characteristics evolution

  • Online:2015-08-31 Published:2015-08-31

摘要: 为定量评估加工工艺对产品可靠性的影响,识别和控制制造过程的薄弱环节,提出基于Granger检验及Cox回归的工艺可靠性综合评估方法。首先对过程特性与产品特性的变量序列进行Granger因果关系检验,构建各道工序的质量特性演化模型,判定加工工艺偏差与工序产品质量偏差之间的定量关系,为建立更准确的评估模型提供依据;结合演化规律,确定影响产品可靠性的工艺过程协变量,建立不带交叉项和带交叉项的Cox比例风险回归模型,通过模型风险比判断过程的关键工序;同时在基准失效率为Weibull分布的假设条件下,给出Cox模型的产品可靠性和平均寿命计算方法;最后对雷达移相单元进行了工艺可靠性综合评估,验证了方法的可行性和有效性。

关键词: 工艺可靠性评估, Granger因果关系检验, 质量特性演化, Cox比例风险模型

Abstract: To evaluate the impact of machining process on product reliability quantitatively,a comprehensive evaluation method for process reliability was put forward based on Granger causality test and Cox Proportional Hazards Regression Model (PHRM).Granger causality test was conducted on the variable series of process characteristics and product characteristics to model the quality characteristics evolution.The quantitative relation between processing deviation and process quality deviation was determined to provide evidence for establishing accurate evaluation model.The process covariate effected on production reliability was confirmed based on evolution rule,and Cox PHRMs with cross term and without cross term were established.Meanwhile,the hazards ratio was given as an index of Cox model to judge the criticalities of different processes.The product reliability and life expectation were calculated under the Weibull Cox PHRM assumption.The case study on process reliability evaluation of radar phase-shift unit was demonstrated to prove the feasibility and effectiveness of the proposed method.

Key words: process reliability evaluation, Granger causality test, quality characteristics evolution, Cox proportional hazards regression model

中图分类号: