Abstract An effective multivariate process control charts for autocorrelated data from multistage processes was established, and the formula for calculating its average run length was given by combining statespace models and statistical process control methods. The correctness of the formula was verified by Monte Carlo simulation. Finally, the new method was illustrated with one real example to demonstrate the implementation steps. The results show that in control the chart has low alarm rate, while out of control it has high alarm rate.
Received: 22 January 2010
Published: 28 September 2010
