Constructing Control Limits for a Sensing System
碩士 === 國立交通大學 === 工業工程與管理系所 === 106 === The sensing system is the main tool used by semiconductor factories to detect the voltage variation of the machine. The sensing system collects multiple data every minute for detection. When the data exceeds the limit set by the system continuously, the system...
Main Authors: | , |
---|---|
Other Authors: | |
Format: | Others |
Language: | zh-TW |
Published: |
2018
|
Online Access: | http://ndltd.ncl.edu.tw/handle/y885hg |
id |
ndltd-TW-106NCTU5031066 |
---|---|
record_format |
oai_dc |
spelling |
ndltd-TW-106NCTU50310662019-05-16T01:24:32Z http://ndltd.ncl.edu.tw/handle/y885hg Constructing Control Limits for a Sensing System 建構感應系統之管制界限 Su, Yi-Ping 蘇怡萍 碩士 國立交通大學 工業工程與管理系所 106 The sensing system is the main tool used by semiconductor factories to detect the voltage variation of the machine. The sensing system collects multiple data every minute for detection. When the data exceeds the limit set by the system continuously, the system will result in a warning signal. It is necessary to stop operating the machine when the system gives a warning signal and find out the reason which causing high voltage of the machine. However, the data distribution of a sensing system is usually follow a non-normal distribution. Therefore the commonly used Shewhart’s control chart will cause false alarms or fail to detect the out-of-control situation for a sensing system. In this case, the voltage variation of the machines becomes difficult to control. When constructing a control chart for non-normal data, the data needs to transform to normal or increase the number of observations for constructing appropriate control limits. However, the data transformation method cannot be always successfully converted to the normal distribution. Increasing sample size will increase the sampling cost. Therefore, the main objective of this study is to apply the Percentile Bootstrap (PB) and Biased-Corrected and Accelerated Percentile Bootstrap (BCa) bootstrap confidence intervals to construct the control limits for a sensing system. The sensitivity analysis is utilized to verify the effectiveness of the proposed method. The results of sensitivity analysis showed that the control limits of the sensing system constructed by the proposed method perform better than that of the traditional methods. Moreover, the control limit constructed by the BCa method out performs PB methods. 唐麗英 李榮貴 2018 學位論文 ; thesis 26 zh-TW |
collection |
NDLTD |
language |
zh-TW |
format |
Others
|
sources |
NDLTD |
description |
碩士 === 國立交通大學 === 工業工程與管理系所 === 106 === The sensing system is the main tool used by semiconductor factories to detect the voltage variation of the machine. The sensing system collects multiple data every minute for detection. When the data exceeds the limit set by the system continuously, the system will result in a warning signal. It is necessary to stop operating the machine when the system gives a warning signal and find out the reason which causing high voltage of the machine. However, the data distribution of a sensing system is usually follow a non-normal distribution. Therefore the commonly used Shewhart’s control chart will cause false alarms or fail to detect the out-of-control situation for a sensing system. In this case, the voltage variation of the machines becomes difficult to control. When constructing a control chart for non-normal data, the data needs to transform to normal or increase the number of observations for constructing appropriate control limits. However, the data transformation method cannot be always successfully converted to the normal distribution. Increasing sample size will increase the sampling cost. Therefore, the main objective of this study is to apply the Percentile Bootstrap (PB) and Biased-Corrected and Accelerated Percentile Bootstrap (BCa) bootstrap confidence intervals to construct the control limits for a sensing system. The sensitivity analysis is utilized to verify the effectiveness of the proposed method. The results of sensitivity analysis showed that the control limits of the sensing system constructed by the proposed method perform better than that of the traditional methods. Moreover, the control limit constructed by the BCa method out performs PB methods.
|
author2 |
唐麗英 |
author_facet |
唐麗英 Su, Yi-Ping 蘇怡萍 |
author |
Su, Yi-Ping 蘇怡萍 |
spellingShingle |
Su, Yi-Ping 蘇怡萍 Constructing Control Limits for a Sensing System |
author_sort |
Su, Yi-Ping |
title |
Constructing Control Limits for a Sensing System |
title_short |
Constructing Control Limits for a Sensing System |
title_full |
Constructing Control Limits for a Sensing System |
title_fullStr |
Constructing Control Limits for a Sensing System |
title_full_unstemmed |
Constructing Control Limits for a Sensing System |
title_sort |
constructing control limits for a sensing system |
publishDate |
2018 |
url |
http://ndltd.ncl.edu.tw/handle/y885hg |
work_keys_str_mv |
AT suyiping constructingcontrollimitsforasensingsystem AT sūyípíng constructingcontrollimitsforasensingsystem AT suyiping jiàngòugǎnyīngxìtǒngzhīguǎnzhìjièxiàn AT sūyípíng jiàngòugǎnyīngxìtǒngzhīguǎnzhìjièxiàn |
_version_ |
1719175808947322880 |