There are three fundamental types of control chart designs namely Shewhart, EWMA, and CUSUM that are used to detect different
kinds of shifts. In this section, firstly, we present the charting structures of these fundamental charts. Later, we will discuss three
selective models to represent different process behaviors including ARMA(p,q), AR(p), and MA(1).
2.1. Shewhart, exponentially weighted moving average, and cumulative sum charts
Control charts are classified regarding to their sensitivity to detect the encountered shift. Firstly, memory-less chart such as Shewhart
chart that is sensitive and more effective in monitoring large shift. It has named memory less because it makes use of the present
information only in constructing the charts. Secondly, memory charts that use both past and present information about the process
in constructing its control chart, which makes it more sensitive and effective to small and moderate shifts in the process. These types
of chart include mainly CUSUM and EWMA. The memory and memory-less charts can be used to monitor and control both location
and dispersion parameters. The details about these main charts are discussed in the coming lines.
2.1.1. Shewhart. It was proposed by Shewhart
15
; it is sensitive for large shifts in the process. Let w be a sample statistic that measures
some quality characteristic of interest with mean μ
w
standard deviation σ
. Based on these parameters, the Shewhart structure in
terms of upper control limit (UCL), center line (CL) and lower control limit (LCL) are defined as