Control charts for binary correlated variables Linda Lee Ho Airlane P Alencar USP - Brazil
Introduction • Automated manufacturing industries • Common practice - Inspections of items produced continuously • n binary random variables, 1 , , Z Z n • ~ Z Bernoulli p i • =1 for a non-conforming item and 0 Z i otherwise
Introduction • Monitoring: the number (or the proportion) of non-conforming items in a sample of n units • np chart or p chart – used for SPC • Assumption: independent Z i
Introduction • However, in many production processes • A common component in the whole process may yield a correlation between the different items. • That is , 0; Corr Z Z i j i j
Introduction • The consequences of this correlation n • is inflated 0 Z i 1 i n • (1 ) Var Z np p (it is overdispersed). i 1 i
Assumptions • New production process • Phase I • For high quality processes: – the correlation of the binary variables is not easily identified. n – High frequency of is expected 0 Z i 1 i
Overdispersed binomial distribution n • T Z the number of non conforming items in the i 1 i sample of correlated n items. n 0 I t n t t t b 1 1 P T t a a b 1 p ; ; ; ln (1 ) a e ln 1 p p b e p • Var(T)=np(1-p)[1+(n-1) ]
Overdispersed binomial distribution
np x np control charts
ML estimation • Maximum likelihood and method of moment estimation of the correlation parameter are presented • Comparison of performance: compared by simulation.
ML Estimation • Likelihood function n k n t t n 1 a b b ( , , , ) 1 1 0 i L T T T a a b i 1 2 k t ( 0, 1,..., ) t i k i i n k t t n nk 0 k a 1 1 i 1 a a b b b 1 i 1 t ( 1.,,. ) i k i
ML Estimation t i ˆ b ˆ ˆ 0 t 1 2 1 ML n n ((1 ) (1 ) ... 1) b b i ˆ ML ML n nk 1 (1 ) b 1 ML k t i i ˆ 1 a ˆ ˆ ML n nk b k b ML ML
ML x MM
np x np control charts H : versus H : H : versus H : . 0 0 1 1 a a 0 1 0 1 • For a type error I equal 5%: ˆ ˆ • Reject null hypothesis if .95 ML | 0 ˆ ˆ • Or a a .05 ML | a 1
Critical values – at 5%
Control chart np • a Shewhart control chart named np to monitor the non-conforming fraction when the binary variables are correlated. • The traditional np chart is a particular case of np the control chart when =0.
np control charts • Upward shifts: p p 1 0 • Given , control limit (CL) is determined
Performance of np , np control charts
Performance of np , np control charts
np x EWMA np control charts
Performance of EWMA np control charts
Conclusions • ML and MM estimation of the correlation parameter are similar. • ML estimator – low bias • the control chart needs at least to double np the sample size - to have the similar performance of the traditional np control chart .
Thank you!
Recommend
More recommend