臺灣博碩士論文加值系統

射出成型是一種常見的塑膠製品製造工藝，廣泛應用於我們日常生活中的各種產品。然而，射出成型設備的昂貴和複雜性使其不適合小規模生產。為了實現高生產量並保持良好的品質，我們需要有效監測射出機的運作狀態。本研究旨在開發一套適用於射出機油泵數據監測的異常檢測系統。其中油泵的震動數據對於本次研究相當重要。一旦震動出現異常，就有可能是設備故障或齒輪軸斷裂等問題，進而影響生產的品質和效率。為了有效的檢測油泵的運作狀態，本研究比較了多種機器學習方法，並採用GridSearchCV、RandomSearchCV和貝葉斯最佳化進一步提升模型性能，且讓模型擁有一定的泛化能力。在感測器上本研究分別使用了整合型感測器、聲發感測器與單軸加速規三種感測器，其中整合型感測器測出的數值較不精確且每個週期的資料量較少，但成本較低，在經過多次的實驗與參數上的調整，也可以做到輔助判斷與展示資料的任務。

Injection molding is a common plastic product manufacturing process and is widely used in various products in our daily lives. However, the expense and complexity of injection molding equipment make it unsuitable for small-scale production. In order to achieve high production throughput and maintain good quality, we need to effectively monitor the operating status of the injection machine. This research aims to develop an anomaly detection system suitable for data monitoring of injection oil pumps. The vibration data of the oil pump is very important for this study. Once the vibration is abnormal, there may be problems such as equipment failure or gear shaft breakage, which will affect the quality and efficiency of production. In order to effectively detect the operating status of the oil pump, this study compared a variety of machine learning methods, and used GridSearchCV, RandomSearchCV and Bayesian optimization to further improve the model performance and give the model a certain generalization ability. In terms of sensors, this study used three types of sensors: integrated sensors, acoustic sensors and single-axis accelerometers. Among them, the values measured by the integrated sensors are less accurate and the values measured in each cycle are The amount of data is small, but the cost is low. After many experiments and parameter adjustments, it can also assist in the task of judging and displaying data.

摘要 I
ABSTRACT II
誌謝 III
目錄 IV
圖目錄 VI
表目錄 VIII
第一章 緒論 1
1.1 研究背景與動機 1
1.1.1 塑膠射出成型機介紹 1
1.1.2 產業需求及挑戰 2
1.2 研究流程 4
1.3 論文架構 5
第二章 文獻探討 6
2.1 射出成型 8
2.2 資料前處理(Data Preprocessing) 10
2.2.1 資料清理(Data Cleaning) 10
2.2.2 資料整合(Data Integration) 11
2.2.3 特徵轉換(Feature Transform) 11
2.3 機器學習模型 14
2.3.1 一類支援向量機(One-Class Support Vector Machine, OCSVM) 15
2.3.2 孤立森林(Isolation Forest, iForest) 18
2.3.3 局部異常因子(Local Outlier Factor, LOF) 21
2.3.4 橢圓包絡(Elliptic Envelope, EE) 23
2.4 超參數最佳化(Hyperparameter Optimization) 25
2.4.1 網格搜索(Grid Search) 25
2.4.2 隨機搜索(Random Search) 27
2.4.3 貝葉斯最佳化(Bayesian Optimization) 29
2.5 評估指標(Evaluation index) 32
2.5.1 混淆矩陣(Confusion matrix) 32
2.5.2 準確率(Accuracy) 34
2.5.3 精確度(Precision) 34
2.5.4 召回率(Recall) 35
2.5.5 F1分數(F1-score) 35
第三章 研究架構與方法 37
3.1 資料蒐集 37
3.2 資料前處理 42
3.2.1 資料清理與資料整合 42
3.2.2 資料特徵轉換 49
3.3 模型訓練與最佳化 54
3.3.1 訓練資料選擇 55
3.3.2 一類支援向量機 57
3.3.3 孤立森林 59
3.3.4 局部異常因子 61
3.3.5 橢圓包絡 63
第四章 研究結果與比較 65
4.1 模型比較 65
4.1.1 一類支援向量機 65
4.1.2 孤立森林 66
4.1.3 局部異常因子 66
4.1.4 橢圓包絡 67
4.2 比較結果 67
第五章 結論與未來展望 69
5.1.1 結論 69
5.1.2 未來展望 69
參考文獻 70

[1]Selvaraj, S. K., Raj, A., Rishikesh Mahadevan, R., Chadha, U., & Paramasivam, V. (2022). A review on machine learning models in injection molding machines. Advances in Materials Science and Engineering, 2022(1), 1949061.
[2]Thiriez, A., & Gutowski, T. (2006, May). An environmental analysis of injection molding. In Proceedings of the 2006 IEEE International Symposium on Electronics and the Environment, 2006. (pp. 195-200). IEEE.
[3]Inasaki, I. (1998). Application of acoustic emission sensor for monitoring machining processes. Ultrasonics, 36(1-5), 273-281.
[4]Liu, J., Zhong, L., Wickramasuriya, J., & Vasudevan, V. (2009, September). User evaluation of lightweight user authentication with a single tri-axis accelerometer. In Proceedings of the 11th International Conference on Human-Computer Interaction with Mobile Devices and Services (pp. 1-10).
[5]Carleo, G., Cirac, I., Cranmer, K., Daudet, L., Schuld, M., Tishby, N., ... & Zdeborová, L. (2019). Machine learning and the physical sciences. Reviews of Modern Physics, 91(4), 045002.
[6]Krenker, A., Bešter, J., & Kos, A. (2011). Introduction to the artificial neural networks. Artificial Neural Networks: Methodological Advances and Biomedical Applications. InTech, 1-18.
[7]Alasadi, S. A., & Bhaya, W. S. (2017). Review of data preprocessing techniques in data mining. Journal of Engineering and Applied Sciences, 12(16), 4102-4107.
[8]Vassiliadis, P. (2009). A survey of extract–transform–load technology. International Journal of Data Warehousing and Mining (IJDWM), 5(3), 1-27.
[9]Chu, X., Ilyas, I. F., Krishnan, S., & Wang, J. (2016, June). Data cleaning: Overview and emerging challenges. In Proceedings of the 2016 international conference on management of data (pp. 2201-2206).
[10]Lenzerini, M. (2002, June). Data integration: A theoretical perspective. In Proceedings of the twenty-first ACM SIGMOD-SIGACT-SIGART symposium on Principles of database systems (pp. 233-246).
[11]Kusiak, A. (2001). Feature transformation methods in data mining. IEEE Transactions on Electronics packaging manufacturing, 24(3), 214-221.
[12]Weisstein, E. W. (2015). Fast fourier transform. https://mathworld. wolfram. com/.
[13]Song, Y. Y., & Ying, L. U. (2015). Decision tree methods: applications for classification and prediction. Shanghai archives of psychiatry, 27(2), 130.
[14]Breiman, L. (2001). Random forests. Machine learning, 45, 5-32.
[15]O'shea, K., & Nash, R. (2015). An introduction to convolutional neural networks. arXiv preprint arXiv:1511.08458.
[16]Chen, T., He, T., Benesty, M., Khotilovich, V., Tang, Y., Cho, H., ... & Zhou, T. (2015). Xgboost: extreme gradient boosting. R package version 0.4-2, 1(4), 1-4.
[17]Schölkopf, B. (1997). Support vector learning (Doctoral dissertation, Oldenbourg München, Germany).
[18]Ito, K., & Kunisch, K. (2008). Lagrange multiplier approach to variational problems and applications. Society for Industrial and Applied Mathematics.
[19]Liu, F. T., Ting, K. M., & Zhou, Z. H. (2008, December). Isolation forest. In 2008 eighth ieee international conference on data mining (pp. 413-422). IEEE.
[20]Kriegel, H. P., Kröger, P., Schubert, E., & Zimek, A. (2009, November). LoOP: local outlier probabilities. In Proceedings of the 18th ACM conference on Information and knowledge management (pp. 1649-1652).
[21]Peña, D., & Prieto, F. J. (2001). Multivariate outlier detection and robust covariance matrix estimation. Technometrics, 43(3), 286-310.
[22]Hubert, M., & Debruyne, M. (2010). Minimum covariance determinant. Wiley interdisciplinary reviews: Computational statistics, 2(1), 36-43.
[23]Yang, L., & Shami, A. (2020). On hyperparameter optimization of machine learning algorithms: Theory and practice. Neurocomputing, 415, 295-316.
[24]Syarif, I., Prugel-Bennett, A., & Wills, G. (2016). SVM parameter optimization using grid search and genetic algorithm to improve classification performance. TELKOMNIKA (Telecommunication Computing Electronics and Control), 14(4), 1502-1509.
[25]Bergstra, J., & Bengio, Y. (2012). Random search for hyper-parameter optimization. Journal of machine learning research, 13(2).
[26]Shahriari, B., Swersky, K., Wang, Z., Adams, R. P., & De Freitas, N. (2015). Taking the human out of the loop: A review of Bayesian optimization. Proceedings of the IEEE, 104(1), 148-175.
[27]Schulz, E., Speekenbrink, M., & Krause, A. (2018). A tutorial on Gaussian process regression: Modelling, exploring, and exploiting functions. Journal of Mathematical Psychology, 85, 1-16.
[28]Luque, A., Carrasco, A., Martín, A., & de Las Heras, A. (2019). The impact of class imbalance in classification performance metrics based on the binary confusion matrix. Pattern Recognition, 91, 216-231.
[29]Swets, J. A. (1988). Measuring the accuracy of diagnostic systems. Science, 240(4857), 1285-1293.
[30]Sokolova, M., Japkowicz, N., & Szpakowicz, S. (2006, December). Beyond accuracy, F-score and ROC: a family of discriminant measures for performance evaluation. In Australasian joint conference on artificial intelligence (pp. 1015-1021). Berlin, Heidelberg: Springer Berlin Heidelberg.
[31]Mahnke, W., Leitner, S. H., & Damm, M. (2009). OPC unified architecture. Springer Science & Business Media.