在人工智慧(artificial intelligence, AI)和物聯網(Internet of Things, IoT)技術的驅動下，智慧製造(smart manufacturing)成為當今熱門的研究議題。製造品質(manufacturing quality, MQ)預測是智慧製造的基礎之一，對於某些無法快速測量品質，或是比較難測量品質的產品，希望可以基於生產前的靜態參數以及生產過程中收集的動態狀態時間序列資料，在產品生產後快速並且準確地預測其製造品質。
本論文提出兩種方法進行線切割放電加工(wire electrical discharge machining, WEDM)工件品質預測，明確地說是進行工件表面粗糙度(surface roughness)預測。第一種方法利用格拉姆角度域(Gramian angular field)轉換時間序列資料為二維圖像，並配合卷積長短期記憶(convolutional long short-term memory, CLSTM)神經網路預測工件表面粗糙度。第二種方法利用馬可夫轉換域(Markov transition field, MTF)轉換時間序列資料為二維圖像，並配合卷積長短期記憶神經網路預測工件表面粗糙度。實驗結果顯示，本論文提出的兩種方法在平均絕對百分比誤差(mean absolute percentage error, MAPE)方面皆優於一個近期提出的相關方法。

Driven by artificial intelligence (AI) and Internet of Things (IoT) technologies, smart manufacturing has become a hot topic today. Predicting manufacturing quality (MQ) is fundamental in smart manufacturing. For some manufactured products whose quality cannot be measured speedily or handily, it is desirable to fast and accurately predict the MQ based on static data, such as manufacturing parameters tuned before production, as well as dynamic data, such as manufacturing conditions gathered during production.
This paper proposes two methods to predict the MQ of wire electrical discharge machining (WEDM), specifically, to predict the workpiece surface roughness (SR). The first method uses Gramian angular field (GAF) to represent dynamic WEDM manufacturing conditions as images, and uses convolutional long short-term memory (CLSTM) neural networks to predict the workpiece SR. The second method uses Markov transition field (MTF) to represent dynamic WEDM manufacturing conditions as images, and uses CLSTM neural networks to predict the workpiece SR. Experiments are conducted to evaluate the performance of the proposed methods. As will be shown, the proposed methods outperform a related method proposed recently in terms of the mean absolute percentage error (MAPE).

[1] K. H. Ho, et al. "State of the Art in Wire Electrical Discharge Machining (WEDM)." International Journal of Machine Tools and Manufacture, Vol. 44, No. 12-13, pp. 1247-1259, Oct. 2004.
[2] K. H. Ho, and S. T. Newman. "State of the Art Electrical Discharge Machining (EDM)." International Journal of Machine Tools and Manufacture, Vol. 43, No. 13, pp.1287-1300, Oct. 2003.
[3] C. L. Fan, and J. R. Jiang. "Surface Roughness Prediction Based on Markov Chain and Deep Neural Network for Wire Electrical Discharge Machining." 2019 IEEE Eurasia Conference on IOT, Communication and Engineering, Oct. 2019.
[4] U. Esme, A. Sagbas, and F. Kahraman. "Prediction of Surface Roughness in Wire Electrical Discharge Machining using Design of Experiments and Neural Networks." Iranian Journal of Science and Technology Transaction B: Engineering, Vol. 33, No. 3, pp. 231-240, June 2009.
[5] A. Kumar, V. Kumar, and J. Kumar. "Prediction of Surface Roughness in Wire Electric Discharge Machining (WEDM) Process based on Response Surface Methodology." International Journal of Engineering and Technology, Vol. 2, No. 4, pp. 708-712, Jan. 2012.
[6] 放電加工原理及應用
https://www.shs.edu.tw/works/essay/2017/03/2017030111003102.pdf, accessed in June 2020.
[7] 線切割放電加工
https://www.coursehero.com/file/52605873/2010-4-e1e12437-7ppt/, accessed in June 2020.
[8] I. Mitiche, et al. "Imaging Time Series for the Classification of EMI Discharge Sources." Sensors, Vol. 18, No. 9, Sep. 2018.
[9] F. R. Sánchez, and J. G. Cervera. "ECG Classification Using Artificial Neural Networks." Journal of Physics: Conference Series, Vol. 1221, No. 1, IOP Publishing, June 2019.
[10] T. A. Nagem, et al. "Deep Learning Technology for Predicting Solar Flares from (Geostationary Operational Environmental Satellite) Data." International Journal of Advanced Computer Science and Applications, Vol. 9, No. 1, Feb. 2018.
[11] C. L. Yang, Z. X. Chen, and C. Y. Yang. "Sensor Classification Using Convolutional Neural Network by Encoding Multivariate Time Series as Two-Dimensional Colored Images." Sensors, Vol. 20, No.1, Dec. 2019.
[12] Neurons introductions:
http://www.hkpe.net/hkdsepe/human_body/neuron.htm, accessed in June 2020.
[13] A. Kaplan, and M. Haenlein. "Siri, Siri, in my hand: Who’s the fairest in the land? On the interpretations, illustrations, and implications of artificial intelligence." Business Horizons, Vol. 62, No. 1, pp. 15-25, Jan. 2019.
[14] Getting started with deep learning:
https://medium.com/@syshen/%E5%85%A5%E9%96%80%E6%B7%B1%E5%BA%A6%E5%AD%B8%E7%BF%92-2-d694cad7d1e5, accessed in June 2020.
[15] Y. LeCun, et al. "Gradient-Based Learning Applied to Document Recognition." Proceedings of the IEEE, Vol. 86, No. 11, pp.2278-2324, Nov. 1998.
[16] S. Hochreiter, and J. Schmidhuber. "Long Short-Term Memory." Neural computation, Vol. 9, No. 8, pp. 1735-1780, Nov. 1997.
[17] Understanding LSTM Networks:
http://colah.github.io/posts/2015-08-Understanding-LSTMs/, accessed in June 2020.
[18] CHMER - Q4025L:
http://www.chmer.com/tw/products-view.php?id=76, accessed in June 2020.
[19] Tokyo Seimitsu – Surfcom 130A:
http://www.accretech.com.cn/surfcom.html, accessed in June 2020.
[20] Surface roughness:
https://tw.misumi-ec.com/pdf/tech/MSM1/Surface_Roughness.pdf, accessed in June 2020.
[21] Z. Wang, and T. Oates. "Encoding Time Series as Images for Visual Inspection and Classification Using Tiled Convolutional Neural Networks." Workshops at the Twenty-Ninth AAAI Conference on Artificial Intelligence, April 2015.
[22] M. Fahim, K. Fraz, and A. Sillitti. "TSI: Time Series to Imaging based Model for Detecting Anomalous Energy Consumption in Smart Buildings." Information Sciences, Vol. 523, pp.1-13, June 2020.
[23] J. R. Jiang. "An improved cyber-physical systems architecture for Industry 4.0 smart factories." Advances in Mechanical Engineering, Vol. 10, No. 6, June 2018.
[24] E. Keogh, et al. "Dimensionality Reduction for Fast Similarity Search in Large Time Series Databases." Knowledge and information Systems, Vol. 3, No. 3, pp.263-286, Aug. 2001.