ch 9 CUSUM and EWMA - 321:QualityControl Chapter9...

Info iconThis preview shows pages 1–15. Sign up to view the full content.

View Full Document Right Arrow Icon
1 321:  Quality Control Chapter 9 Cumulative Sum and Exponentially Weighted  Moving Average Control Charts Instructor :  Linda Boyle Industrial and Systems Engineering  University of Washington
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
2 Introduction   To   detect large abrupt shifts (1.5 σ  to 2 σ  shifts  or larger) Shewhart Control Charts. To detect small shifts CUSUM: CUSUM:  cumulative sum  control chart  EWMA: EWMA:  exponentially weighted moving  average
Background image of page 2
3 Example Process mean =10 std dev=1 Process mean =11 std dev=1 Subgroup x Subgroup x 1 9.5   9 11.2 2 7.0 10 10.3 3 9.3 11 11.1 4 11.7 12 10.4 5 12.2 13 11.6 6 10.2 14 11.9 7 8.0 15 10.9 8 11.5
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
4 Example 5 6 7 8 9 10 11 12 13 14 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 UCL(new) UCL(orig) CL(new) CL(orig) LCL(new) LCL(orig)
Background image of page 4
5 CUSUM chart CUSUM chart Incorporates all information in the sequence  of sample values  by plotting the  cumulative sums  of the deviations  of the sample values from a target value. Where C i  = cumulative sum up to and including  the i th  sample and n> 1 μ - = = i 1 j 0 j i ) x ( C
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
6 CUSUM Using previous  data Where  μ o =10 μ - = = i 1 j 0 j i ) x ( C Subgroup x x 1 9.5 9 11.2 2 7.0 10 10.3 3 9.3 11 11.1 4 11.7 12 10.4 5 12.2 13 11.6 6 10.2 14 11.9 7 8.0 15 10.9 8 11.5 Board discussion
Background image of page 6
7 CUSUM Subgroup x i C i Subgroup x i C i 1 9.5 -0.5 9 11.2 0.6 2 7.0 -3.5 10 10.3 0.9 3 9.3 -4.2 11 11.1 2.0 4 11.7 -2.5 12 10.4 2.4 5 12.2 -0.3 13 11.6 4.0 6 10.2 -0.1 14 11.9 5.9 7 8.0 -2.1 15 10.9 6.8 8 11.5 -0.6 Plot C i
Background image of page 7

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
8 CUSUM -6.0 -4.0 -2.0 0.0 2.0 4.0 6.0 8.0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Sample number C i μ = 10 μ = 11
Background image of page 8
9 CUSUM Plot Is not a control chart Lacks statistical control limits Two types of CUSUM representation Tabular cusum (preferred method) Can add Fast Initial Response (FIR) or Headstart V-mask Error probability consideration,  α β The book strongly advises against using it (see pg  416)
Background image of page 9

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
10 Tabular Cusum Let x i  be the i th  observation on the process If the process is in control then x i  has a  normal distribution Assume  σ  is known or can be estimated Upper cusum: Lower cusum: C i + = max 0, x i - ( m 0 + K ) + C i - 1 + [ ] C i - = max 0,( m 0 - K ) - x i + C i - 1 - [ ]
Background image of page 10
11 Tabular Cusum Upper Cusum The accumulated change from the target  μ 0   above  the target with one statistic, C + Lower Cusum The accumulated change from the target  μ 0   below  the target with another statistic, C
Background image of page 11

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
12 Tabular Cusum starting values are If either statistic exceed a decision interval  H , the process is  considered to be out of control. Often taken as  H  = 5 σ C i + = max 0, x i - ( m 0 + K ) + C i - 1 + [ ] C i - = max 0,( m 0 - K ) - x i + C i - 1 - [ ] 0 0 0 = = - + C C Where K = reference value (or allowance or slack value )
Background image of page 12
13 Tabular Cusum   Selecting  K K  is often chosen as halfway between target  μ 0  and,  out-of-control value we are interested in detecting (mean  μ 1 Shift is expressed in standard deviation units as  μ 1 = μ 0 +δσ , then  K  is: 2 2 K 0 1 μ - μ = σ δ =
Background image of page 13

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
14 Tabular Cusum Example   Subgroup
Background image of page 14
Image of page 15
This is the end of the preview. Sign up to access the rest of the document.

{[ snackBarMessage ]}

Page1 / 49

ch 9 CUSUM and EWMA - 321:QualityControl Chapter9...

This preview shows document pages 1 - 15. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online