Table 1.
Control chart constants for R chart of normal order statistics based on different sampling schemes.
Table 2.
Control chart constants for S chart of normal order statistics based on different sampling schemes.
Table 3.
RL properties of the proposed CUSUM—R chart for detecting increases in process standard deviation (n = 5, ARL0 = 200).
Table 4.
RL properties of the proposed CUSUM—S chart for detecting increases in process standard deviation (n = 5, ARL0 = 200).
Table 5.
RL properties of the proposed CUSUM—R chart for detecting decreases in process standard deviation (n = 5, ARL0 = 200).
Table 6.
RL properties of the proposed CUSUM—S chart for detecting decreases in process standard deviation (n = 5, ARL0 = 200).
Table 7.
RL comparison among CUSUM charts for monitoring increases in standard deviation (n = 5, ARL0 = 500, σopt = 1.3).
Table 8.
ARL comparison among CUSUM charts for monitoring decreases in standard deviation (n = 5, ARL0 = 500, σopt = 0.7).
Fig 1.
ARL curves for imperfect scheme I versus CUSUM-S2, CUSUM-ln S2, EWMA-S2 and EWMA-ln S2 charts for detecting 1.2σ when n = 5 and ARL0 = 200.
Fig 2.
ARL curves for imperfect scheme II versus CUSUM-S2, CUSUM-ln S2, EWMA-S2 and EWMA-ln S2 charts for detecting 1.2σ when n = 5 and ARL0 = 200.
Fig 3.
ARL curves for imperfect scheme III versus CUSUM-S2, CUSUM-ln S2, EWMA-S2 and EWMA-ln S2 charts for detecting 1.2σ when n = 5 and ARL0 = 200.
Fig 4.
Standard deviations of the thirty re-sampled datasets collected using SRS, schemes I, II and III.
Fig 5.
CUSUM S control chart using classical and proposed scheme I with n = 5.
Fig 6.
CUSUM S control chart using classical and proposed scheme II with n = 5.
Fig 7.
CUSUM S control chart using classical and proposed scheme III with n = 5.
Fig 8.
CUSUM S control chart using classical and imperfect scheme I with n = 5.
Fig 9.
CUSUM S control chart using classical and imperfect scheme II with n = 5.
Fig 10.
CUSUM S control chart using classical and imperfect scheme III with n = 5.