臺灣博碩士論文加值系統

統計製程管制為工業製程重要的監控技術，而又以管制圖為最常用的手法，為了幫助從業人員監控管制圖樣式迅速找出造成製程異常的癥結點並加以改善，許多關於管制圖樣式辨識的研究先後被提出。協助製程工程師找出製程中的可歸屬原因，改進製程上的不良變異。
目前大多數的模擬研究皆假設管制圖樣式符合常態分配；但實務上的製程也很有可能為非常態分配。為了協助非常態管制圖樣式辨識，本研究以支撐向量機（Support Vector Machines ,SVM）作為管制圖樣式辨識系統並結合小波分析與基因演算法，藉此提高樣式辨識的準確度和辨識速度，並使用正確辨識率（rate of correct classification ,ROCC）作為本研究評估指標。
模擬後發現以結合小波分析與基因演算法的支撐向量機辨識系統辨識非常態管制圖樣式可以得到良好的正確辨識率，且峰度的大小會對辨識率造成些許的影響，而偏度的不同則對辨識率無明顯差異。在模擬後比較前處理辨識及參數搜索方法後，小波分析及基因演算法皆能獲得較高的辨識率且更為穩定。

Statistical process control is an important technology for monitoring industrial processes, and the control chart is the most commonly used method. In order to help practitioners monitor the control charts pattern, quickly identify and improve the symptoms of process abnormalities.

At present, most simulation studies assume that the control chart style conforms to the normal distribution However, the process is also to be an non-normal distribution. Assist process engineers to identify the cause of the process and improve the process variation. In order to assist the identification of the non-normal control chart pattern, this study uses Support Vector Machines （SVM） as the control chart pattern recognition system and combines wavelet analysis and gene algorithm to improve the accuracy and recognition speed of pattern recognition. The rate of correct classification （ROCC） was used as the evaluation index of this study.

After simulation, it is found that the support vector machine identification system combined with wavelet analysis and gene algorithm can obtain a good correct classification rate, and the kurtosis will have a slight influence on the recognition rate, but the skewness is different. There is no significant difference in the recognition rate. After comparing the pre-processing identification and parameter search methods after simulation, both wavelet analysis and gene algorithm can obtain higher recognition rate and be more stable.

摘要 i
Abstract ii
目錄 iii
表目錄 v
圖目錄 vi
第一章、緒論 1
1.1 研究背景與動機 1
1.2 研究目的 3
1.3 論文架構 4
第二章、文獻探討 5
2.1 支撐向量機（Support Vector Machine, SVM） 5
2.1.1 支撐向量機的理論 5
2.1.2 線性支撐向量機 6
2.1.3 非線性支撐向量機 7
2.1.4 建構支撐向量機分類模型 8
2.2 管制圖的樣式辨識 10
2.2.1 傳統管制圖 10
2.2.2 管制圖樣式辨識 10
2.3 小波分析於管制圖樣式辨識 12
2.3.1 小波分析簡介 13
2.3.2 Haar小波轉換 13
2.3.3 連續小波轉換 14
2.3.4 離散小波轉換 14
2.4 結合基因演算法於支撐向量機 16
2.4.1 基因演算法 16
2.4.2 基因演算法在支撐向量機的應用 17
2.5 支撐向量機於非常態管制圖樣式辨識之研究 17
第三章、研究方法 19
3.1 研究流程 19
3.2 訓練樣本產生方法 20
3.2.1 蒙地卡羅模擬法 20
3.2.2 非常態分配 20
3.2.3 管制圖樣式 22
3.3 特徵擷取方法 23
3.3.1 小波轉換 23
3.3.2 統計特徵值 25
3.4 基因演算法搜索參數 25
3.4.1 染色體設計 26
3.4.2 適應度函數（Fitness function） 26
3.5 建構支撐向量機分類模型 27
3.6 辨識視窗 27
3.7 效益評估 28
第四章、研究結果與分析 29
4.1 模擬流程 29
4.2 非常態分配下的辨識系統績效 32
4.3 前處理辨識比較 35
4.4 參數搜索辨識比較 36
第五章、結論與建議 40
5.1 結論 40
5.2 建議 41
參考文獻 42
附錄 44

1. Akande, K. O., Owolabi, T. O., Twaha, S., Olatunji, S. O. (2014). Performance Comparison of SVM and ANN in Predicting Compressive Strength of Concrete. Journal of Computer Engineering, 16:88-94.
2. Bag, M., Gauri, S. K., & Chakraborty, S. (2012). An expert system for control chart pattern recognition. The International Journal of Advanced Manufacturing Technology, 62:291–301.
3. Bakas, J., Mahalat, M., & Mollah, A. (2016). A Comparative Study of Various Classifiers for Character Recognition on Multi-script Databases. International Journal of Computer Applications, 155:1-5.
4. Bradley, J. V. (1968). Distribution-Free Statistical Tests. Prentice Hall, New Jersey.
5. Chen, Z., Lu, S., Lam, S. (2007). A hybrid system for SPC concurrent pattern recognition. Advanced Engineering Informatics, 21:303-310.
6. Cheng, H. P., & Cheng, C. S. (2009). Control chart pattern recognition using wavelet analysis and neural networks. 品質學報, 16(5), 311-321.
7. Chinnam, R. B. (2002). Support vector machines for recognizing shifts in correlated and other manufacturing processes. International Journal of Production Research, 40:4449-4466.
8. Cortes, C., Vapnik, V. (1995). Support-Vector Network. Machine Learning, 20:273-297.
9. Cuentas, S., Niebles, R, P., Garcia, E. (2017). Support vector machine in statistical process monitoring: a methodological and analytical review. The International Journal of Advanced Manufacturing Technology, 91:485–500
10. Du, S., Huang, D., Lv, J. (2013). Recognition of concurrent control chart patterns using wavelet transform decomposition and multiclass support vector machines. Computers & Industrial Engineering, 66:683-695.
11. Guh, R. S. (2002). Effects of non-normality on artificial neural network based control chart pattern recognizer. Journal of the Chinese Institute of Industrial Engineers, 19:13-22.
12. Hachicha, W., Ghorbel, Ahmed. (2012). A survey of control-chart pattern-recognition literature (1991–2010) based on a new conceptual classification scheme. Computers & Industrial Engineering, 63 :204–222.
13. Heisele, B., Ho, P., Poggio, T. (2001). Face recognition with support vector machines: global versus component-based approach. Proceedings Eighth IEEE International Conference on Computer Vision, 2:688-694.
14. Holland, J. H. (1992). Adaptation in natural and artificial systems, MIT Press, Cambridge.
15. Huang, C. L., Wang, C. J. (2006). A GA-based feature selection and parameters optimization for support vector machines. Expert Systems with Applications, 31:231-240.
16. Lu, C. J., Shao, Y. E., Li. C. C. (2014）. Recognition of Concurrent Control Chart Patterns by Integrating ICA and SVM. Applied Mathematics & Information Sciences, 8:681-689.
17. Nelson, L. S. (1985). Interpreting Shewhart X control charts. J. Qual. Technol, 17(2):114–116.
18. Ranaee, V., Ebrahimzadeh, A. (2011). Control chart pattern recognition using a novel hybrid intelligent method. Applied Soft Computing, 11:2676-2686.