在射出成型加工製程中，機台加工參數偏移會使產品品質特性與所期望之不同，而加工參數往往眾多且複雜，故常需花費大量的時間與人力成本於異常排除的工作上。本研究主要探討多品質射出成型機台加工參數之異常診斷方法，採用聚乳酸/玻璃纖維複合材料作為試驗樣本，射出成型機台加工參數包括熔膠溫度、射出速度、保壓壓力、保壓時間與冷卻時間，試片品質包含拉伸強度、衝擊強度、硬度與彎曲強度。實驗方面則透過調整機台加工參數使其偏離最佳加工條件，使其成為異常加工參數之樣本，並利用多變量統計管制圖對樣本品質作管制。以最佳加工條件所得之樣本當作實驗的歷史數據，Hotelling’s T2找出最佳加工參數組合之管制上限，並由管制上限檢測出異常樣本之T2值，再透過殘差管制法針對異常T2值的樣本取得拉伸、衝擊、硬度和彎曲四項品質殘差值，可藉由殘差值建構出一倒傳遞類神經網路分類器，透過此網路辨識異常的加工參數，或根據變異數分析找出品質與加工參數間關係，建立一單品質異常診斷表快速地找出異常加工參數。實驗結果證明本研究所提出的診斷方法，對於辨識各異常加工參數可以有良好的診斷效果，並可以建立一套射出成型加工參數異常的診斷系統。

In the injection molding processing, complex processing parameters need to be adjusted for expected qualities. Once the process is abnormal, it’s essential to spend a lot of time and human work in fault diagnosis. In this study, we focus on fault diagnosis of injection molding processing parameters for polylactide/glass fiber composites. The injection molding processing parameters include melt temperature, injection speed, packing pressure, packing time, and cooling time. The qualities include tensile strength, hardness, impact strength and flexure strength. In this thesis, we adjust the processing parameters to make the process conditions deviate from the optimal process condition, and the multivariate statistical control chart can be used to monitor downgrade qualities. The machine is operated at the optimal process conditions to generate normal samples and their four qualities data are chosen as historical data. With these historical data, the upper control limit(UCL) of optimal processing parameters can be found by using Hotelling’s T2, and is used to detect the fault T2 values with abnormal samples. Then, we obtain the residuals of qualities for abnormal samples by residual control chart, and choose them to be the feature values for the neural network in distinguishing fault processing parameters. On the other hand, we build a fault diagnosis table for single quality based on the relationship between the quality and significant processing parameters by using analysis of variance in Taguchi method. The results show that the proposed methods can diagnose fault processing parameters at a high rate.

1. 蕭基昊，聚乳酸/玻璃纖維複合材料於射出成型之多品質加工參數最佳化研究，台灣科技大學材料科學與工程學系碩士論文(2015)。
2. A. K. Bledzki, A. Jaszkiewicz and D. Scherzer, “Mechanical properties of PLA composites with man-made cellulose and abaca fibers,” Composites: Part A, Vol. 40, pp. 404-412, 2009.
3. A. Jaszkiewicz, A. K. Bledzki and P. Franciszczak, “Improving the mechanical performance of PLA composites with natural, man-made cellulose and glass fibers - a comparison to PP counterparts,” Polimery Journal, Vol. 58, No. 6, pp. 435-442, 2013.
4. M. H. Chung, “Study on weld-line of ABS thin-wall injection molding parts,” Master Thesis, Department of Mechanical Engineering, Chung Yuan Christian University, 2001.
5. Y. K. Yang, J. R. Shie, H. T. Liao, J. L. Wen, and R. T. Yang, "A study of Taguchi and design of experiments method in injection molding process for polypropylene components," Journal of Reinforced Plastics and Composites, Vol. 27, No. 8, pp. 819-834, 2008.
6. N. H. M. Nar, N. Muhamad, M. H. I. Ibrahim, M. Ruzi and K. R. Jamaludin, "Optimization of injection molding parameter of Ti-6Al-4V powder mix with palm stearin and polyethylene for the highest green strength by using Taguchi method," International Journal of Mechanical an Materials Engineering, Vol. 6, No,1, pp. 126-132, 2011.
7. N. M. Mehat and S. Kamarddin, "Investigating the effects of injection molding parameters on the mechanical properties of recycled plastic parts using the Taguchi method," Materials and Manufacturing Process, Vol. 26, No. 2, pp. 202-209, 2011.
8. E. Kuram, E. Tasci, A. L. Altan, M. M. Medar, F. Yilmaz, and B. Ozcelik, "Investigating the effects of recycling number and injection parameters on the mechanical properties of glass-fiber reinforced nylon 6 using Taguchi method," Materials and Design, Vol. 49, pp. 139-150, 2013.
9. H. Hotelling, “Multivariate quality control-illustrated by air testing of sample bombsights,” Techniques of Statistical Analysis, Mcgraw-Hill, New York, pp. 11-184, 1947.
10. D. M. Hawkins, “Multivariate quality control based on regression-adjusted variables,” Technometrics, Vol. 33, pp. 61-75, 1991.
11. 王亞倫，診斷多變量管制圖之研究，國立中央大學工業管理研究所碩士論文(2000)。
12. H. H. Yang, M. L. Huang, and S. W. Yang, “Integrating auto-associative neural networks with hotelling’s T2 control charts for wind turbine fault detection” Energies, Vol. 8, No. 10, pp. 12100-12115, 2015.
13. L. H. Chen and T. Y. Wang, “Artificial neural networks to classify mean shifts from multivariate χ2 chart signals”, Computers and Industrial Engineering, Vol. 47, pp. 195-205, 2004.
14. S. Zeng, T. Chen, L. Wang, and H. Qu, “Monitoring batch-to-batch reproducibility using direct analysis in real time mass spectrometry and multivariate analysis: A case study on precipritation,” Journal of Pharmaceutical and Biomedical Analysis, Vol. 76, pp. 87-95, 2013.
15. B. Igne, Z. Shi, J. K. Drennen 3rd, and C. A. Anderson, “Effect and detection of raw material variability on the performance of near-infrared calibration models for pharmaceutical products,” Journal of Pharmaceutical Sciences, Vol. 103, No. 2, pp. 545-556, 2014.
16. Z. Wang, Y. Liu and P. J. Griffin. "Neural net and expert system diagnose transformer faults," IEEE Computer Application in Power, Vol. 13, pp. 50-55, 2000
17. C. S. Cheng, “A multi-layer neural network model for detecting changes in the process mean”, Computers and Industrial Engineering, Vol. 28, pp.51-61, 1995.
18. G. F. Bin, J. J. Gao, X. J. Li, and B. S. Dhillon, “Early fault diagnosis of rotating machinery based on wavelet packets-empirical mode decomposition feature extraction and neural network,” Mechanical Systems and Signal Processing, Vol. 27, pp. 696-711, 2012.
19. P. K. Kankar, S. C. Sharma, and S. P. Harsha, “Fault diagnosis of ball bearing using machine learning methods,” Expert Systems with Applications, Vol. 38, No. 3, pp. 1876-1886, 2011.
20. 葉怡成，類神經網路模式應用與實作，儒林圖書有限公司(2000)。
21. 羅華強，類神經網路：MATLAB的應用，高立圖書(2011)。
22. R. L. Mason and J. C. Young, “Improving the sensitivity of the T2 statistic in multivariate process control,” Journal of Quality Technology, Vol. 31, No. 2, pp. 155-165, 1999.
23. D. C. Montgomery, Statistical Quality Control, 5th ed., John Wiley and Sons Inc(2005).
24. 李德治，林孟儒，童惠玲，統計學第二版，博碩文化(2012)。