這項研究旨在實現一套機械手臂研磨診斷系統，透過機械手臂原生的電控箱數據與少量的感測器訊號就能達到砂帶磨耗程度變化的監視與工件表面品質的預測。本研究以機械手臂夾取鋁塊對砂帶進行研磨實驗，過程中收取機械手臂電控箱的扭矩訊號、桌上型砂輪機的電流訊號、以及裝設於機械手臂夾爪上的加速規訊號，在初步的資料切分階段對收取到的訊號進行研磨狀態與非研磨狀態的資料分割，接續經過資料前處理後再輸入人工智慧模型中。此外，砂帶使用白光干涉儀照射後以白光干涉儀配套軟體計算出的參數作為評判磨耗等級的標準。本研究使用多層感知機分類模型來預測砂帶磨耗等級，也使用多層感知機回歸模型來預測工件表面粗糙度，經由測試結果得出，在預測砂帶磨耗等級時，使用機械手臂訊號預測準確率可達58%，加上研磨機電流後可達69%，再加上加速規訊號後預測準確率可達84%，在預測工件表面粗糙度時增加電流特徵與加速規特徵均方誤差皆可達0.03以下。

This study aims to accomplish a robotic grinding diagnostic system, which can monitor the abrasive belt wear level and predict the surface quality of the workpiece through the robotic arm's controlling box data and few sensor signals. In this study, the robot arm is used to grasp the aluminum workpiece for abrasive belt grinding experiment. During the process, the torque signals from the controlling box of the robot arm, the current signals from the grinding machine, and the acceleration signals from the clamping jaws of the robot arm were collected. In addition, the abrasive belt is measured by the white light interferometer and the parameters calculated by the white light interferometer software are used to evaluate the abrasion level. In this study, a multi-layer perceptron classification model was used to predict the wear level of the abrasive belt, and a multi-layer perceptron regression model was used to predict the surface roughness of the workpiece. When predicting the abrasive belt wear level, the prediction accuracy can reach 58% with the robotic arm’s torque signal, 69% after adding the current signal of the grinding machine, and 84% after adding both current signal and acceleration signal. When predicting the surface roughness of the workpiece, adding the sensor signal can slightly improve the prediction accuracy, and the mean square error was less than 0.03. When predicting the abrasive belt wear parameters, the sensor signal signature input can reduce the prediction error to about 0.02.

1. International Federation of Robotics, 2020, "Executive Summary World Robotics 2020 Industrial Robots",
https://ifr.org/img/worldrobotics/Executive_Summary_WR_2020_Industrial_Robots_1.pdf
2. Iglesias, I., Sebastián, M.A., Ares, J.E., 2015, "Overview of the state of robotic machining: Current situation and future potential", Proc. Engineering, 132, pp. 911-917.
3. Ji, W., Wang, L., 2019, "Industrial robotic machining: a review", J. Int. Adv. Manuf. Technol , 103(1-4), pp 1239-1255.
4. Jinno , M., 1995, "Development of a force controlled robot for grinding, chamfering and polishing", J. Int. Conf. Robot. Autom., 2, pp. 1455–1460.
5. Zhu, D., Feng, X., Xu, X., Yang, Z., Li, W., Yan, S., Ding, H. , 2020,"Robotic grinding of complex components:A step forward efficient and intelligent machining – challenges,solutions, and applications", Robotics and Computer-Integrated Manufacturing, 65, Paper Number 101908
6. Zhu, D., Feng, X., Xu, X., Yang, Z., Li, W., Yan, S., Ding, H., 2018, "Analysis and assessment of robotic belt grinding mechanisms by force modelingand force control experiments", J. Int. Tribology, 120, pp. 93-98.
7. Xie, X.,Sun, L, 2016, "Force control based robotic grinding system and application", Proc. WCICA, pp 2552-2555.
8. Ma, K., Wang, X., Shen,D., 2018, "Design and experiment of robotic belt grinding system with constant grinding force", Proc. M2VIP
9. He, Z.,Li, J,Liu, Y, Yan, J., 2020, "Single-grain cutting based modeling of abrasive belt wear in cylindrical grinding",Friction, 8(1), pp 208-220.
10. Xiao, G., Huang, Y., 2017, "Experimental research and modelling of life-cycle material removal in belt finishing for titanium alloy", J. Manufacturing Processes, 30, pp 255-267.
11. Jia, F., Lei, Y.,Guo, L.,Lin J.,Xing,S., 2018, "A neural network constructed by deep learning technique and its application to intelligent fault diagnosis of machines", Neauralcomputing, 272, pp 619-628.
12. Liu, H., Zhou J., Xu, Y., Zheng Y., Peng, X., Jiang, W., 2018, "Unsupervised fault diagnosis of rolling bearings using a deep neural network based on generative adversarial networks", Neauralcomputing, 315, pp 412-424.
13. Chen, Z.,Gryllias, K., Li, W., 2019, "Mechanical fault diagnosis using Convolutional Neural Networks and Extreme Learning Machine", Mechanical System and Signal Processing, 133
14. Cao, X., C., Chen, B., Q., Yao, B., He, W., P., 2019, "Combining translation-invariant wavelet frames and convolutional neural network for intelligent tool wear state identification", Computers in Industrial, 106, pp 71-84.
15. Islam, M., M., M., Kim, J., M., 2019, "Automated bearing fault diagnosis scheme using 2D representation ofwavelet packet transform and deep convolutional neural network", Computers in Industrial, 106, pp 142-153.
16. López de Calle, K., Ferreiro, S., Arnaiz, A., Sierra, B., 2019, "Dynamic condition monitoring method based on dimensionality reduction techniques for data-limited industrial environments", Computers in Industrial, 112
17. Thomazella, R., Lopes, W., N., Aguiar, P., R., Alexandre, F., A., Fiocchi, A., A., Bianchi, E., C., 2019, "Digital signal precessing for self-vibration monitoring in grinding : A new approach based on the time-frequency analysis of vibration signals", J. int. Measurement Confederation, 145, pp 71-83.
18. Arun, A., Rameshkumar, K., Unnikrishnan, D., Sumesh, A., 2018, "Tool condition monitoring of cylindrical grinding using acoustic emission sensor", Proc. Materials Today, 5, pp 11888-11899.
19. Nguyen, D., Yin, S., Tang, Q., Son, P., X., Duc, L., A., 2019, "Online monitoring of surface roughness and grinding wheel wear when grinding Ti-6Al-4V titanium alloy using ANFIS-GPR hybrid algorithm and Taguchi analysis", Precision Engineering, 55, pp 275-292.
20. Tanaka, K., Koshy, P., 2019, "A pneumatic sensor for grinding wheel condition monitoring", Precision Engineering, 56, pp 62-68.
21. Xu, L., Niu, M., Zhao, D.,Xing, N., Fan, F., 2019, "Methodology for the immediate detection and treatment of wheel wear in contour grinding", Precision Engineering, 60, pp 405-412.
22. Fukuhara, Y., Suzuki, S., Sasahara, H., 2018, "Real-time grinding state discrimination strategy by use of monitor-embedded grinding wheels", Precision Engineering, 51, pp 128-136.
23. Sachin, K. ,P., Rameshkumar, K. , 2020, "Grinding wheel condition prediction with discrete hidden Markov model using acoustic emission signature. ", Materials Today: Proceedings.
24. Hase, A., Mishan, H., Wada, M., 2013, "Microscopic study on the relationship between AE signal and wear amount", Wear, 308(1-2), pp 142-147.
25. Qu, X., Zeng, P., Xu, C., Fu, D., 2017, "RNN-based method for fault diagnosis of grinding system", J. Int. Conf. CYBER, pp 673-678.
26. Cheng ,C., Li, J. Y., Liu ,Y. M., Nie, M., Wang, W. X., 2019, "Deep convolutional neural network-based in-process tool condition monitoring in abrasive belt grinding", Computers in Industry, 106, pp 1-13.
27. Cheng ,C., Li, J. Y., Liu ,Y. M., Nie, M., Wang, W. X., 2020, "An online belt wear monitoring method for abrasive belt grinding under varying grinding parameters" , J. Manufacturing Processes, 50, pp 80-89.
28. Zhang, X., Chen, H., Xu, J., Song, X., Wang, J., Chen, X., 2018, "A novel sound-based belt condition monitoring method for robotic grindingusing optimally pruned extreme learning machine", J. Materials Processing Technology, 260, pp 9-19.
29. Serope, K.,Steven, R. S., 2008, Manufacturing Process for Engineering Materials, Pearson Education, US.
30. Anderson, D.,Warkentin, A.,Bauer, R., 2011, "Experimental and numerical investigations of single abrasive-grain cutting", J. Int. Machine Tools and Manufacture, 51(12), pp 898-910.
31. Baidakova, N., V., Orlova, T., N., 2017, "Influence of abrasive grain geometrical characteristics on the grinding quality", Procedia Engineering, 206, pp 194-199.
32. Pandiyan,V. , Tijahjowidodo, T , ASamy Meena, P., 2016, "In process surface roughness estimation model For compliant abrasive belt machining process", Proc. CIRP, 46, pp 254-257.