Change Parameters And Multivariate Quality Control Charts,

Quality control chart is a useful tool for monitoring produce process in statistics state and reducing waster. Adaptive control charts and multivariate control chart possess important significance. In this paper, I study above two issues and gain some results as follows:1.Under population follow normal distribution hypothesis, designing combine median-range adaptive control charts with variable sampling interval, variable sample size and variable control limit .It may be regarded as previously median-range extension. Data shows that such chart is much superior than any other chart.2.Using Johnson curve distribution transformation method, describing nonnomal data to normal data or approximate normal data. At the same time, giving an example for illuminating this method.3.Discussing application of entropy in multivariate control chart, Firstly, testing whether the correlation structure has changed by means of the mutual information approach. Secondly, using approximately approach to discuss two estimators of conditional entropy. Finally, establishing corresponding entropy chart.