质量特性非正态分布规律的研究

经典的统计过程控制理论基于质量特性服从正态分布，应用于当前高精度质量控制，会因非正态分布质量特性而产生误差。针对生产现场质量特性测量数据，分析并从理论上通过检验证实了分布的非正态性，提出以Beta分布拟合非正态分布质量特性，给出了拟合方法，检验了其合理性，为准确地对非正态分布质量特性实施过程控制提供分布模型。

Study on the Non-normal Distribution Rule of Quality Characteristic

The classical theory of the statistic process control bases on the hypothesis that quality characteristic obeys the normal distribution, which applied in the current high-precision control will bring inaccuracy or even mistake because of the quality characteristic＇s no-normal distribution. This paper has analyzed and validated the no-normal distribution, using the test for the data measured from the production, advocated using beta distribution to describe it, put forward the method and tested the reasonableness. The conclusion shows that it is a fit distribution model for statistic process control theory to control the non-normal QC accurately.