机械加工过程的输出分布和性能分析

Output Distribution and Capability Analysis of Mechanical Manufacturing Processes

在机械加工过程中,由于设备和环境等因素的影响,一些过程输出中除随机误差以外,还存在着不可消除的系统因素导致的过程输出的系统性变差或趋势变化,出现周期性停工或调整的情况。此类过程的输出分布特征随系统因素的不同而不同,传统的统计过程控制方法的控制效果可能失效,特别是对于过程性能评价。本文以两种常见情况为例,讨论了非时变和时变系统因素产生的加工系统变差与随机变差的合成分布的特征,给出了合成分布的标准概率密度函数(pdf)。在此基础上,提出了过程输出具有此类分布特征的过程性能分析及合格率估计的新方法。最后指出了此类合成分布对过程调整和优化的作用。

There are system deviations or trend drifts from non-removable special causes besides random deviations in process outputs because of the changes in equipment and environment, resulting in periodic shutdown or adjustment. In these eases the statistical process control method may fail, especially for process performce assessment. The compound distribution of system deviations and random deviations is discussed through two familiar eases in practice. One is a non-time-variety ease and another is of time-variety. The standard probability density functions of the distribution is presented in the paper. Based on these functions, new methods for assessing process performance and calculating its rate of pass in mechanical manufacturing processes are developed. Finally the paper points out the effects of the compound distribution on the optimization and adjustment of processes.