Process Variance Monitoring Based On Cumulative Sum Control Chart

Product quality control in production process,it is usually the monitoring of the process mean,and the monitoring of the process variance is ignored.For small deviations in the process,people are increasingly aware of the importance of monitoring process variance.The thesis studies the cumulative sum method of process variance monitoring,and combines the cumulative sum method with the conventional control limit idea,which effectively improves the sensitivity of monitoring process variance when there is a small deviation.Firstly,the paper proposes the upper one-sided Shewhart-cusum S2 chart,discusses the design ideas,and gives the corresponding upper one-sided conventional control limit coefficients for different sample sizes of n,and the average running chain length is compared with the average running length of the traditional variance control chart,and the influence of sample size on its monitoring efficiency is studied,which effectively improves the sensitivity of monitoring the upward shift of process variance.Secondly,the paper studies the lower one-sided Shewhart-cusum S2 chart design,gives the corresponding unilateral regular control limit coefficients for different sample sizes n,and analyzes the effect of sample size on its monitoring efficiency influences.In comparison with the monitoring performance of the traditional variance control chart,it is proved that the monitoring efficiency of the downward shift of the process variance is significantly improved by the lower one-sided Shewhart-cusum S2 chart.Finally,the paper proposes a two-sided Shewhart-cusum S2 chart design for the situation of non-directional deviation of variance,and gives the corresponding two-sided conventional control limit coefficients for different sample sizes n.Compared with the traditional Shewhart S2 chart,the monitoring efficiency of the two-sided Shewhart-cusum S2 chart is higher.When the sample size becomes larger,the monitoring sensitivity of the two-sided Shewhart-cusum S2 chart is improved.