Weighted Likelihood Ratio Test-Based Control Chart For Monitoring The Process Variability

Some factors in manufacturing such as false raw material, lower-skilled operators, or loosening of machine setting maybe lead to a change in process dispersion without necessarily impacting on the level of the process mean. In this paper, we propose a new chart with weighted likelihood ratio test(WLRT) statistics for detecting the process variance. The new chart can be easily designed and constructed. Computation results imply that it supplies a satisfactory performance, including the monitor of the increase and the decrease in variability and the individual observation at the sampling point which are universal in many practical applications. In this paper, run length(RL) distributions for upper-sided and low-sided WLRT charts are described. The optimal parameters which can be used as a design aid in choosing specific parameter values based on the average run length(ARL)are described.The application of our proposed approach is illustrated by a real example of chemical process control.