The Improvement About The Shewhart Control Charts

Shewhart control charts in quality management has became an indispensablemanagement tool. It is a kind of chart with control limits. It can distinguish the causeof quality fluctuation by accidental or systems, in order to determine the productionprocess in control. The papers analysis Shewhart control chart summed up that have abetter monitor in a large shift for the process mean, but it is not enough sensitive in asmall shift. so that the signal can not be investigated to find out the cause andeliminate them. In order to better play the role of control charts, the paper discussedthe Shewhart control chart improvements.The main contents are as follows:Firstly, thesis describes the sequential method in the course of the development,research and the basics at home and abroad. Then introduces several different qualitycontrol charts.Secondly, we proposed an improved median chart based on Klein model of xcontrol chart. The comparison with the Shewhart x showed that this chart was a littlesuperior in performance for small process average drift, This paper further ameliorateits out of control condition on the basis of previous research. Obtain the twoconsecutive samples median functions, using the finite integrals illustrations, andusing approximating method to calculate the average chain length. Using theapproximation method to calculate the average run length. it is numerically verifiedthat the sumx chart works more effectively than we mentioned the tradition Shewhartx chart and rrx chart,especially, in the process with slight shift.Again, we propose a control chart called VGR control chart,which is acombination of Shewhart’s X chart and an extended version of conforming runlength chart and variable sample. It is numerically verified that, the VGR chart givesa reduction in out of control ATS as compared to those of the GR control chart and Xchart when in-control ATS is not smaller than a certain number.Steady stateperformance of the VGR is also better. Hence, VGR chart turns out to be economicalto detect small shifts in the process mean. Finally, we use numerical solution method and structure the function aboutcontrol limits,calculate the minimum with fminsearch of MATLAB software.optimization of the upper and lower control limits.The results show that after definedthe average run length precisely, given the expression of the calculated average runlength of the CCC-r, and thus eliminating the ARL-biased phenomenon of CCC-rcontrol chart, It can effectively reduce the reaction time of the CCC-r control chartsfor monitoring the increasing nonconforming rate, enhanced the sensitivity ofmonitoring degradation of products.