全自動虛擬量測 (Automatic Virtual Metrology, AVM) 系統已成功地應用於半導體等高科技產業，其可將具有量測延遲特性之抽檢改為線上且即時之全檢，而半導體業之製程特性為大量且近於穩態。由於精密機械廠對於產品工件的全檢精度要求持續提高，若要應用AVM系統於工具機產業，以達到即時且全自動的產品全檢之標的時，其挑戰在於，如何調整AVM的預測與決策機制，以能因應少量且具半穩態製程特性的精密加工產業。
本研究提出目標值調整機制 (Target-Value Adjustment Scheme, TVA Scheme) 與自動化取樣決策機制 (Automated Sampling Decision Scheme, ASD Scheme)。TVA可自動調整AVM的預測目標值，以因應相同建模用於不同類型之生產條件，包含不同加工尺寸與公差範圍等。而ASD則可動態地調整AVM所需之抽測率，在維持一定水準之預測精度下降低所需之量測成本。
實際加工研究案例顯示，在應用於汽車鋁輪圈加工時，TVA可因應不同加工條件，減少更新預測模型所需樣本數，可從與建模端相同樣本數，改善至僅需二到三筆樣本即能維持平均預測精度 (MAE) 在0.01mm內；而ASD則可在標準件加工之量產且穩定製程環境下，降低所需抽測率從100%改善為7.57%。此外，上述技術亦應用於航太產業中的飛機引擎機匣加工之精度預測，其不僅實現了重要精度線上即時全檢之目標，也達成了能降低量測成本之目的。

Automatic Virtual Metrology (AVM) system has been successfully applied to many high-tech industries such as the semiconductor industry. It can convert sampling inspection with metrology delay into real-time and online total inspection. Currently, the precision machinery factories’ demands for enhancing the workpiece prediction accuracy are continuously increasing. The major challenge of applying the AVM system to the machine tool industry in order to achieve online and real-time total inspection is to adjust the AVM prediction and decision-making schemes from steady and mass production as in the semiconductor industry to small volume and semi-steady production in the precision machine tool industry.
This study proposes the Target-Value Adjustment Scheme (TVA Scheme) and Automated Sampling Decision Scheme (ASD Scheme). TVA can adjust the target values automatically to cope with the issue of applying the same model creation to various types of workpiece machining conditions, including different machining dimensions and tolerance ranges. ASD can dynamically adjust the sampling rates that AVM requires to reduce the measurement cost while still maintaining good prediction accuracy.
The actual machining case studies show that after applying TVA to the wheel machining automation, TVA can reduce the sample count required for model-refreshing in response to different machining conditions. TVA can also reduce the need of refreshing the same sample count as for model creation to only 2-3 samples, which are sufficient for maintaining good prediction accuracy.
As for ASD, under the stable mass-production environment, such as standard workpieces machining, it can reduce the sampling rate from 100% to 7.57% at best. The techniques described above are also applied to the aerospace industry in the prediction accuracy of the aircraft engine casing machining. TVA and ASD not only help to realize the goal of online and real-time total inspection, but also are verified to effectively reduce the measurement cost.

[1] F.-T. Cheng, C.-A. Kao, C.-F. Chen, and W.-H. Tsai, “Tutorial on Applying the VM Technology for TFT-LCD Manufacturing,” IEEE Transactions on Semiconductor Manufacturing, vol. 28, no. 1, pp. 55-69, 2015. doi:10.1109/TSM.2014.2380433.
[2] F.-T. Cheng, H.-C. Huang, and C.-A. Kao, “Developing an Automatic Virtual Metrology System,” IEEE Trans. on Automation Science and Engineering, vol. 9, no. 1, pp. 181-188., Jan. 2012.
[3] F.-T. Cheng, H.-C. Huang, and C.-A. Kao, “Dual-Phase Virtual Metrology Scheme,” IEEE Trans. on Semiconductor Manufacturing, vol. 20, no. 4, pp. 566-571, Nov. 2007.
[4] G. Valiño, C. M. Suárez, J. C. Rico, B. J. Álvarez, and, D. Blanco, “Comparison between a Laser Micrometer and a Touch Trigger Probe for Workpiece Measurement on a CNC Lathe,” Advanced Materials Research, vol. 498, pp. 49-54, 2012. doi: 10.4028/www.scientific.net/AMR.498.49.
[5] H.-C. Yang, H. Tieng, and F.-T. Cheng, “Automatic Virtual Metrology for Wheel Machining Automation,” to appear in International Journal of Production Research. doi: 10.1080/00207543.2015.1109724. (has been published online)
[6] H. Tieng, H.-C. Yang, and F.-T. Cheng, “Total Precision Inspection of Machine Tools with Virtual Metrology.” in Proceedings of the 2015 IEEE International Conference on Automation Science and Engineering (CASE2015), Gothenburg, Sweden, pp. 1446-1447, August 24-28, 2015.
[7] F.-T. Cheng, Y.-T. Chen, Y.-C. Su, and D.-L. Zeng, “Evaluating Reliance Level of a Virtual Metrology System,” IEEE Transactions on Semiconductor Manufacturing, vol. 21, no. 1, pp. 92-103, February 2008.
[8] Y.-T. Huang and F.-T. Cheng, “Automatic Data Quality Evaluation for the AVM System,” IEEE Transactions on Semiconductor Manufacturing, vol. 24, no. 3, pp.445-454, August 2011.
[9] 許智誠「適用於全自動虛擬量測系統的新穎量測資料品質評估架構」，國立成功大學製造資訊與系統研究所碩士論文學位論文，2014
[10] F.-T. Cheng, C.-F. Chen, Y.-S. Hsieh, H.-H. Huang, and C.-C. Wu, “Intelligent Sampling Decision Scheme Based on the AVM System,” International Journal of Production Research, vol. 53, no. 7, pp. 2073-2088, 2015. doi:10.1080/00207543.2014.955924.
[11] F.-T. Cheng, Y.-S. Hsieh, C.-F. Chen, and J.-R. Lyu, “Automated Sampling Decision Scheme Based on the AVM System,” International Journal of Production Research, doi: Aug. 2015. doi: 10.1080/00207543.2015.1072649.
[12] R. Teti, K. Jemielniak, G. O’Donnell, and D. Dornfeld, “Advanced Monitoring of Machining Operations,” CIRP Annals-Manufacturing Technology, vol. 59, no. 2, pp. 717-739, 20112. doi:10.1016/j.cirp.2010.05.010.
[13] J. V. Abellan-Nebot and F. R. Subirón, “A Review of Machining Monitoring Systems based on Artificial Intelligence Process Models.” The International Journal of Advanced Manufacturing Technology, vol. 47, no. 1-4, pp. 237-257, 2010. doi:10.1007/s00170-009-2191-8.
[14] M. Chandrasekaran, , M. Muralidhar, C. M. Krishna, and U. S. Dixit, “Application of Soft Computing Techniques in Machining Performance Prediction and Optimization: a Literature Review,” The International Journal of Advanced Manufacturing Technology, vo. 46 no. 5-8, pp. 445-464, 2010. doi:10.1007/s00170-009-2104-x.
[15] Y.-T. Huang and F.-T. Cheng, “Automatic Data Quality Evaluation for the AVM System.” IEEE Transactions on Semiconductor Manufacturing, vol. 24, no. 3, pp. 445-454, 2011. doi:10.1109/TSM.2011.2154910.
[16] J. Mou and C. R. Liu, “A Method for Enhancing the Accuracy of CNC Machine Tools for On-Machine Inspection,” Journal of Manufacturing Systems, vol. 11, no. 4, pp. 229-237, 1992. doi:10.1016/0278-6125(92)90023-9.
[17] N. Ghosh , Y. B. Ravi, A. Patra, S. Mukhopadhyay, S. Paul, A. R. Mohanty, and A. B. Chattopadhyay, “Estimation of Tool Wear during CNC Milling Using Neural Network-Based Sensor Fusion.” Mechanical Systems and Signal Processing, vol. 21, no. 1, pp. 466-479, 2005. doi:10.1016/j.ymssp.2005.10.010.
[18] R. E. Haber, J. E. Jiménez, C. R. Peres, and J. R. Alique, “An Investigation of Tool-Wear Monitoring in a High-Speed Machining Process.” Sensors and Actuators A: Physical, vol. 116, no. 3, pp. 539-545, 2004. doi:10.1016/j.sna.2004.05.017.
[19] 黃暄恆「適用於全自動虛擬量測系統之智慧型參數篩選機制」，國立成功大學製造資訊與系統研究所碩士論文學位論文，2013.