國立虎尾科技大學 |

Language: English

Back to Search results for [ subject:"Back propagation." ]

結合製程統計特徵值與類神經網路於管制圖異常形狀之辨識 = Improve...

顧瑞祥

結合製程統計特徵值與類神經網路於管制圖異常形狀之辨識 = Improved neural network-based control chart pattern recognition using raw data and statistical data simultaneously

Record Type:	Language materials, printed : monographic
Paralel Title:	Improved neural network-based control chart pattern recognition using raw data and statistical data simultaneously
Author:	施炳光,
Secondary Intellectual Responsibility:	顧瑞祥,
Secondary Intellectual Responsibility:	國立虎尾科技大學
Place of Publication:	雲林縣
Published:	國立虎尾科技大學;
Year of Publication:	民96[2007]
Edition:	初版
Description:	118面圖,表 : 30公分;
Subject:	倒傳遞類神經網路
Subject:	Back Propagation Neural Network
Online resource:	http://140.130.12.251/ETD-db/ETD-search-c/view_etd?URN=etd-0522107-170641
Summary:	管制圖可以用來決定系統的狀態並偵測製程中隨時可能發生的異常情況。異常的管制圖形狀與製程變異中一些特殊的非機遇性原因有關聯，因此有效地辨識異常管制圖形狀能減少可能需要的檢查次數，並加速診斷搜尋。近年來類神經網路已經成功地應用在管制圖形狀辨識上，但大多數的研究均以原始資料作為類神經網路的輸入向量(Raw data based, RB)；部分研究則是利用由原始資料所擷取出的特徵當作輸入向量(Feature data based, FB)以減少網路規模。本論文中所使用的訓練範例與測試範例均為利用蒙地卡羅模擬法產生出生產線製程數據並結合由製程數據所擷取出的統計特徵值一同當作倒傳遞類神經網路的輸入向量(Hybrid data based, HB)，再利用類神經網路軟體來訓練與偵測異常管制圖形狀。本論文同時討論在常態環境與自相關環境下使用HB與RB於靜態下對異常管制圖形狀辨識做績效測試，並再對HB做動態測試模擬。靜態測試結果得知在常態環境下RB的平均辨識率為92.99％，HB的平均辨識率為95.89％，而在自相關環境下RB的平均辨識率為92.11％，HB的平均辨識率為95.12％，代表在常態環境下使用HB能擁有較佳的辨識能力，而在自相關環境下也仍然維持著良好的辨識績效。此外，由動態測試結果得知在常態環境下HB的平均辨識率為87％，而在自相關環境下HB的平均辨識率則為82％，兩者數據均明顯比靜態測試時來的要差，而自相關的動態測試則較常態環境下略差，但辨識率還可維持在80％以上，代表著使用HB在常態環境下擁有不錯的辨識能力，而在自相關環境也仍然維持著一定的辨識績效。 Control chart patterns (CCPs) can be used to determine the status of system. Unnatural CCPs can be associated with a particular set of assignable causes for process variation. In recent years, artificial neural networks (ANNs) have been successfully used in the CCP recognition task. In intelligent SPC, most of researches used raw data (RB) as input vector and the other researches have used statistical feature data extracted from raw data (FB) as input vector for reducing network size. In this thesis, we present an ANN-based approach, in which an improved hybrid training data (HB) integrates both the time series data (Raw data) and the statistical feature data (Feature data). The training data set and testing data set used in this thesis were generated by Monte-Carlo Simulation Method for production line process data. Both HB and RB will be examined in normal environment and auto-correlated environments at a static state while performing the tests of abnormal CCP recognition and then simulating HB at a dynamic state. The static test result shows that the average recognition rates of RB and HB are 92.99

結合製程統計特徵值與類神經網路於管制圖異常形狀之辨識 = Improved neural network-based control chart pattern recognition using raw data and statistical data simultaneously
施, 炳光

結合製程統計特徵值與類神經網路於管制圖異常形狀之辨識 = Improved neural network-based control chart pattern recognition using raw data and statistical data simultaneously / 施炳光撰 - 初版. - 雲林縣 : 國立虎尾科技大學, 民96[2007]. - 118面 ; 圖,表 ; 30公分.
倒傳遞類神經網路Back Propagation Neural Network

顧, 瑞祥

結合製程統計特徵值與類神經網路於管制圖異常形狀之辨識 = Improved neural network-based control chart pattern recognition using raw data and statistical data simultaneously
LDR:04165nam0 2200253 450 001 540262
010 0 $b 平裝
100 $a 20090417h akaa0chia50020302ba
101 0 $a chi
102 $a cw
105 $a ak am 000yy
200 1 $a 結合製程統計特徵值與類神經網路於管制圖異常形狀之辨識 $d Improved neural network-based control chart pattern recognition using raw data and statistical data simultaneously $f 施炳光撰
205 $a 初版
210 $a 雲林縣 $d 民96[2007] $c 國立虎尾科技大學
215 0 $a 118面 $c 圖,表 $d 30公分
314 $a 指導教授：顧瑞祥
328 $a 碩士論文--國立虎尾科技大學工業工程與管理究所
330 $a 管制圖可以用來決定系統的狀態並偵測製程中隨時可能發生的異常情況。異常的管制圖形狀與製程變異中一些特殊的非機遇性原因有關聯，因此有效地辨識異常管制圖形狀能減少可能需要的檢查次數，並加速診斷搜尋。近年來類神經網路已經成功地應用在管制圖形狀辨識上，但大多數的研究均以原始資料作為類神經網路的輸入向量(Raw data based, RB)；部分研究則是利用由原始資料所擷取出的特徵當作輸入向量(Feature data based, FB)以減少網路規模。本論文中所使用的訓練範例與測試範例均為利用蒙地卡羅模擬法產生出生產線製程數據並結合由製程數據所擷取出的統計特徵值一同當作倒傳遞類神經網路的輸入向量(Hybrid data based, HB)，再利用類神經網路軟體來訓練與偵測異常管制圖形狀。本論文同時討論在常態環境與自相關環境下使用HB與RB於靜態下對異常管制圖形狀辨識做績效測試，並再對HB做動態測試模擬。靜態測試結果得知在常態環境下RB的平均辨識率為92.99％，HB的平均辨識率為95.89％，而在自相關環境下RB的平均辨識率為92.11％，HB的平均辨識率為95.12％，代表在常態環境下使用HB能擁有較佳的辨識能力，而在自相關環境下也仍然維持著良好的辨識績效。此外，由動態測試結果得知在常態環境下HB的平均辨識率為87％，而在自相關環境下HB的平均辨識率則為82％，兩者數據均明顯比靜態測試時來的要差，而自相關的動態測試則較常態環境下略差，但辨識率還可維持在80％以上，代表著使用HB在常態環境下擁有不錯的辨識能力，而在自相關環境也仍然維持著一定的辨識績效。 Control chart patterns (CCPs) can be used to determine the status of system. Unnatural CCPs can be associated with a particular set of assignable causes for process variation. In recent years, artificial neural networks (ANNs) have been successfully used in the CCP recognition task. In intelligent SPC, most of researches used raw data (RB) as input vector and the other researches have used statistical feature data extracted from raw data (FB) as input vector for reducing network size. In this thesis, we present an ANN-based approach, in which an improved hybrid training data (HB) integrates both the time series data (Raw data) and the statistical feature data (Feature data). The training data set and testing data set used in this thesis were generated by Monte-Carlo Simulation Method for production line process data. Both HB and RB will be examined in normal environment and auto-correlated environments at a static state while performing the tests of abnormal CCP recognition and then simulating HB at a dynamic state. The static test result shows that the average recognition rates of RB and HB are 92.99
510 1 $a Improved neural network-based control chart pattern recognition using raw data and statistical data simultaneously
610 0 $a 倒傳遞類神經網路 $a 管制圖 $a 蒙地卡羅模擬法 $a 製程統計特徵值
610 1 $a Back Propagation Neural Network $a Control chart pattern $a Monte-Carlo Simulation Method $a statistical feature
681 $a 008.169M $b 0899
700 $a 施 $b 炳光 $3 523743
702 $a 顧 $b 瑞祥 $3 541107
712 $a 國立虎尾科技大學 $b 工業工程與管理究所 $3 523742
770 $a Ping-Kuang Shih $3 586681
772 $a Ruey-Shiang Guh $3 586682
801 0 $a cw $b 虎尾科技大學 $c 20071207 $g CCR
801 2 $a cw $b 虎尾科技大學 $c 20090417 $g CCR
856 7 $u http://140.130.12.251/ETD-db/ETD-search-c/view_etd?URN=etd-0522107-170641