受益於全球AI浪潮以及蓬勃發展的智慧製造需求，機械手臂在產業中越發普及，所看重的即為機械手臂的高靈活度以及能應用在多項複雜且精細的工作任務。有鑑於此，機械手臂的精準度至關重要，然而現行提升機械手臂精度方法需仰賴較昂貴的量測儀器，不利於現場人員使用。因此本論文主要目的為研究具有高便利性、低成本且高精度之運動學參數鑑別方法以補償幾何誤差。本論文根據不同量測儀器推導相對應之參數鑑別模型，最後採用具有低成本之基於治具之運動學參數鑑別模型。本論文所提出基於治具之兩個運動學參數鑑別模型，係利用治具間的相對位置作為量測值並以治具間的角度以及長度作為約束條件，採用幾何誤差校正演算法對運動學參數進行鑑別，並以電腦模擬驗證其能有效提昇機械手臂在笛卡爾空間之精度。在實際上機時，本論文針對所提出之運動學參數鑑別模型進行驗證，並使用不同幾何校正演算法，檢驗是否能有效降低長度以及角度約束條件，實驗結果顯示本論文之方法能夠有效補償運動學幾何學參數誤差。最後引入深度增強式學習，期望對校正後的運動學參數進一步補償其殘餘誤差。

In response to the global wave of artificial intelligence (AI) and the growing demand for smart manufacturing, robotic arms have become more popular in industry. The popularity comes from their high flexibility and capable of performing a wide variety of complex and precise tasks, making positional accuracy a crucial factor. Therefore, the accuracy of robotic arms is indispensable importance. Existing methods for improving robotic accuracy often rely on expensive measurement equipment, which are not suitable for on-site personnel. Therefore, this thesis aims to develop a kinematic parameter identification method that offer high convenience, low cost, and high accuracy for compensation geometric error. Based on different types of measurement equipment, corresponding parameter identification models are derived. Ultimately, a low-cost kinematic parameter identification model based on a fixture is adopted for implementation. Two fixture-based calibration models are proposed, using the relative positions and lengths between fixtures as constraints. Different geometric error correction algorithm is employed, using a regression-based parameter identification method to obtain the calibrated kinematic parameters. Simulation results demonstrate that the proposed method effectively improves the accuracy of the robotic arm in Cartesian space. The proposed kinematic parameter identification model is further verified through real-world robotic implementation. Different geometric calibration algorithm are applied to examine whether the length and angle constraint errors can be effectively reduced. The results indicate that the proposed method effectively compensates for kinematic geometric parameter errors. Finally, deep reinforcement learning is introduced to further compensate for the residual errors of the kinematic parameters after calibration.

[1] 2024/2025產業技術白皮書，經濟部技術處，2024
[2] “IFR 發表 2025 年機器人五大發展趨勢，AI、人型、永續性成重要關鍵”， Accessed: Jun 02, 2025. [Online]. Available: https://technews.tw/2025/02/10/ifr-2025-report
[3] “機器人精度提升校正技術，機械工業雜誌”，Accessed: Jun 02, 2025. [Online].Available:https://www.automan.tw/tw/magazine-period/202408-497/2935
[4] A. Y. Elatta, L. P. Gen, F. L. Zhi, D. Yu, and F. Luo, “An overview of robot calibration,” Information Technology Journal, vol. 3, no. 1, pp. 74–78, 2004.
[5] Y. Wu, A. Klimchik, S. Caro, B. Furet, and A. Pashkevich, “Geometric calibration of industrial robots using enhanced partial pose measurements and design of experiments,” Robotics and Computer-Integrated Manufacturing, vol. 35, pp. 151–168, Feb. 2015.
[6] Z. Wang, Y. Liu, X. Wang, D. Zhang, L. Qin, and J. Liu, “Field calibration method for industrial robots based on single position sensitive device,” IEEE Transactions on Instrumentation and Measurement, vol. 72, pp. 1–12, 2023.
[7] T. Chen, S. Li, Y. Qiao, and X. Luo, “A robust and efficient ensemble of diversified evolutionary computing algorithms for accurate robot calibration,” IEEE Transactions on Instrumentation and Measurement, vol. 73, pp. 1–14, 2024.
[8] Y. Gan, J. Duan, and X. Dai, “A calibration method of robot kinematic parameters by drawstring displacement sensor,” International Journal of Advanced Robotic Systems, vol. 16, no. 5, pp. 1–12, Oct. 2019.
[9] S. Hayati and M. Mirmirani, “Improving the absolute positioning accuracy of robot manipulators,” Journal of Robotic Systems, vol. 2, no. 4, pp. 397–413, 1985.
[10] H. Zhuang, Z. S. Roth, and F. Hamano, “A complete and parametrically continuous kinematic model for robot manipulators,” in Proc. IEEE International Conference on Robotics and Automation (ICRA), Cincinnati, OH, USA, May 1990.
[11] T. Sun, C. Liu, B. Lian, P. Wang, and Y. Song, “Calibration for precision kinematic control of an articulated serial robot,” IEEE Transactions on Industrial Electronics, vol. 68, no. 7, pp. 6000–6009, Jul. 2020.
[12] X. Chen and Q. Zhan, “The kinematic calibration of an industrial robot with an improved beetle swarm optimization algorithm,” IEEE Robotics and Automation Letters, vol. 7, no. 2, pp. 4694–4701, Apr. 2022.
[13] Z. Jiang, W. Zhou, H. Li, Y. Mo, W. Ni, and Q. Huang, “A new kind of accurate calibration method for robotic kinematic parameters based on the extended Kalman and particle filter algorithm,” IEEE Transactions on Industrial Electronics, vol. 65, no. 4, pp. 3337–3345, Apr. 2018.
[14] M. Vincze, J. P. Prenninger, and H. Gander, “A laser tracking system to measure position and orientation of robot end effectors under motion,” The International Journal of Robotics Research, vol. 13, no. 4, pp. 305–314, Aug. 1994.
[15] G. Luo, L. Zou, Z. Wang, C. Lv, J. Ou, and Y. Huang, “A novel kinematic parameters calibration method for industrial robot based on Levenberg–Marquardt and differential evolution hybrid algorithm,” Robotics and Computer-Integrated Manufacturing, vol. 71, pp. 102165, 2021.
[16] W. Yang, S. Li, Z. Li, and X. Luo, “Highly accurate manipulator calibration via extended Kalman filter-incorporated residual neural network,” IEEE Transactions on Industrial Informatics, vol. 19, no. 11, pp. 10831–10841, Nov. 2023.
[17] Z. Li, S. Li, O. O. Bamasag, A. Alhothali, and X. Luo, “Diversified regularization enhanced training for effective manipulator calibration,” IEEE Transactions on Neural Networks and Learning Systems, vol. 34, no. 11, pp. 8778–8790, Nov. 2023.
[18] H. Chen, S. Zhang, G. Wang, Y. Gao, M. Sun, W. Lu, H. Song, and Z. Shi, “LASSO based compensation method for geometric errors of large coordinate measuring machine,” Measurement, vol. 196, pp. 111157, Jun. 2022.
[19] N. A. Barakat, M. A. Elbestawi, and A. D. Spence, “Kinematic and geometric error compensation of a coordinate measuring machine,” International Journal of Machine Tools and Manufacture, vol. 40, no. 6, pp. 833–850, 2000.
[20] L. Žlajpah and T. Petrič, “Kinematic calibration for collaborative robots on a mobile platform using motion capture system,” Robotics and Computer-Integrated Manufacturing, vol. 79, pp. 102446, 2023.
[21] H. Stone, A. Sanderson, and C. Neuman, “Arm signature identification,” in Proc. 1986 IEEE International Conference on Robotics and Automation (ICRA), vol. 3, San Francisco, CA, USA, Apr. 1986, pp. 41–48.
[22] H. W. Stone and A. C. Sanderson, “Statistical performance evaluation of the S-model arm signature identification technique,” in Proc. 1988 IEEE International Conference on Robotics and Automation (ICRA), Philadelphia, PA, USA, Apr. 1988, pp. 939–946.
[23] Z. Jiang, M. Huang, X. Tang, B. Song, and Y. Guo, “Observability index optimization of robot calibration based on multiple identification spaces,” Autonomous Robots, vol. 44, pp. 1029–1046, 2020.
[24] G. Du, Y. Liang, C. Li, P. X. Liu, and D. Li, “Online robot kinematic calibration using hybrid filter with multiple sensors,” IEEE Transactions on Instrumentation and Measurement, vol. 69, no. 9, pp. 7092–7107, Sep. 2020.
[25] 魯昶甫，工業用機械手臂運動學參數鑑別與補償研究，碩士論文，國立成功大學，電機工程學系，2019
[26] A. Nubiola and I. A. Bonev, “Absolute calibration of an ABB IRB 1600 robot using a laser tracker,” Robotics and Computer-Integrated Manufacturing, vol. 29, no. 1, pp. 236–245, Feb. 2013.
[27] M. Meloun, J. Militký, K. Kupka, and R. G. Brereton, “The effect of influential data, model and method on the precision of univariate calibration,” Talanta, vol. 57, no. 4, pp. 721–740, 2002.
[28] Y. Liu, Y. Li, Z. Zhuang, and T. Song, “Improvement of robot accuracy with an optical tracking system,” Sensors, vol. 20, no. 21, pp. 6341, Nov. 2020.
[29] I. W. Park, B. J. Lee, S. H. Cho, Y. D. Hong, and J. H. Kim, “Laser-based kinematic calibration of robot manipulator using differential kinematics,” IEEE/ASME Transactions on Mechatronics, vol. 17, no. 6, pp. 1059–1067, Dec. 2011.
[30] Y. Liu, Z. Zhuang, and Y. Li, “Closed-loop kinematic calibration of robots using a six-point measuring device,” IEEE Transactions on Instrumentation and Measurement, vol. 71, pp. 1–12, 2022.
[31] G. Gao, Y. Li, F. Liu, and S. Han, “Kinematic calibration of industrial robots based on distance information using a hybrid identification method,” Complexity, vol. 2021, no. 1, pp. 8874226, 2021.
[32] A. Omodei, G. Legnani, and R. Adamini, “Three methodologies for the calibration of industrial manipulators: Experimental results on a SCARA robot,” Journal of Robotic Systems, vol. 17, no. 6, pp. 291–307, 2000.
[33] W. Xu, D. Li, and M. Wang, “Complete calibration of industrial robot with limited parameters and neural network,” in Proceedings of the 2016 IEEE International Symposium on Robotics and Intelligent Sensors (IRIS), Tokyo, Japan, Dec. 2016, pp. 103–108.
[34] D. Chen, T. Wang, P. Yuan, N. Sun, and H. Tang, “A positional error compensation method for industrial robots combining error similarity and radial basis function neural network,” Measurement Science and Technology, vol. 30, no. 12, pp. 125010, 2019.
[35] H. Zhong, C. Hu, X. Li, L. Gao, B. Zeng, and H. Dong, “Kinematic calibration method for a two-segment hydraulic leg based on an improved whale swarm algorithm,” Robotics and Computer-Integrated Manufacturing, vol. 59, pp. 361–372, 2019.
[36] Z. Li, S. Li, and X. Luo, “Using quadratic interpolated beetle antennae search to enhance robot arm calibration accuracy,” IEEE Robotics and Automation Letters, vol. 7, no. 4, pp. 12046–12053, Oct. 2022.
[37] C. Gong, J. Yuan, and J. Ni, “Nongeometric error identification and compensation for robotic system by inverse calibration,” International Journal of Machine Tools and Manufacture, vol. 40, no. 14, pp. 2119–2137, 2000.
[38] D. K. Bilal, M. Unel, L. T. Tunc, and B. Gonul, “Development of a vision based pose estimation system for robotic machining and improving its accuracy using LSTM neural networks and sparse regression,” Robotics and Computer-Integrated Manufacturing, vol. 74, pp. 102262, 2022.
[39] G. Zhao, P. Zhang, G. Ma, and W. Xiao, “System identification of the nonlinear residual errors of an industrial robot using massive measurements,” Robotics and Computer-Integrated Manufacturing, vol. 59, pp. 104–114, 2019.
[40] H.-N. Nguyen, J. Zhou, and H.-J. Kang, “A calibration method for enhancing robot accuracy through integration of an extended Kalman filter algorithm and an artificial neural network,” Neurocomputing, vol. 151, pp. 996–1005, Mar. 2015.
[41] Z. Wang, Z. Chen, Y. Wang, C. Mao, and Q. Hang, “A robot calibration method based on joint angle division and an artificial neural network,” Mathematical Problems in Engineering, vol. 2019, no. 1, pp. 9293484, 2019.
[42] F. Meng, J. Wei, Q. Feng, Z. Dong, R. Kang, D. Guo, and J. Yang, “A robot error prediction and compensation method using joint weights optimization within configuration space,” Applied Sciences, vol. 14, no. 24, pp. 1–20, 2024.
[43] A.C. Reddy, “Difference between Denavit-Hartenberg (D-H) classical and modified conventions for forward kinematics of robots with case study,” International Conference on Advanced Materials and manufacturing Technologies, 2014, pp. 267-286.
[44] P. M. Kebria, S. Al-Wais, H. Abdi, and S. Nahavandi, ‘‘Kinematic and dynamic modelling of UR5 manipulator,’’ in Proc. IEEE International Conference on Systems, Man, and Cybernetics, Oct. 2016, pp. 4229–4234.
[45] K. S. Fu, R. C. Gonzalez, C. G. Lee, and H. Freeman, Robotics: Control, Sensing, Vision, and Intelligence, vol. 1. New York, NY, USA: McGraw-Hill, 1987.
[46] R. Keating and N. J. Cowan, "UR5 Inverse Kinematics," unpublished, Johns Hopkins University, 2016. Available: https://tianyusong.com/wp-content/uploads/2017/12/ur5_inverse_kinematics.pdf.
[47] B. W. Mooring, Z. S. Roth, and M. R. Driels, Fundamentals of Manipulator Calibration. Hoboken, NJ, USA: Wiley, 1991.
[48] S. Xu, X. Jia, J. Liu, and T. Li, “Hierarchical compensation of robot positioning error: Addressing geometric and nongeometric influences,” IEEE Transactions on Instrumentation and Measurement, vol. 74, pp. 1–15, 2025.
[49] J. J. Moré, “The Levenberg–Marquardt algorithm: Implementation and theory,” in Numerical Analysis, G. A. Watson, Ed. Cham, Switzerland: Springer, 1978, pp. 105–116.
[50] R. Ge, S. M. Kakade, R. Kidambi, and P. Netrapalli, “The step decay schedule: A near optimal, geometrically decaying learning rate procedure for least squares,” in Advances in Neural Information Processing Systems, vol. 32, Dec. 2019
[51] S. Fujimoto, H. Hoof, and D. Meger, “Addressing function approximation error in actor-critic methods,” in Proc. of 35th International Conference on Machine Learning (ICML), Stockholm, Sweden, Jul. 2018, pp. 1582–1591.
[52] R. S. Sutton and A. G. Barto, Reinforcement Learning: An Introduction. Cambridge, MA, USA: MIT Press, 1998.
[53] P. Sigron, I. Aschwanden, and M. Bambach, “Compensation of geometric, backlash, and thermal drift errors using a universal industrial robot model,” IEEE Transactions on Automation Science and Engineering, vol. 21, no. 4, pp. 6615–6627, Oct. 2024.
[54] Y. Wang, Z. Chen, H. Zu, X. Zhang, C. Mao, and Z. Wang, “Improvement of heavy load robot positioning accuracy by combining a model-based identification for geometric parameters and an optimized neural network for the compensation of nongeometric errors,” Complexity, vol. 2020, Art. no. 5896813, Jan. 2020.
[55] W. K. Veitschegger and C. H. Wu, “Robot calibration and compensation,” IEEE Journal of Robotics and Automation, vol. 4, no. 6, pp. 643–656, Dec. 1988.
[56] E. K. Chong and S. H. Żak, An Introduction to Optimization. Hoboken, NJ, USA: John Wiley & Sons, 2013.
[57] J. Zhao, X. Zhang, and J. Zhao, “A Levenberg–Marquardt method for tensor approximation,” Symmetry, vol. 15, no. 3, pp. 694, Mar. 2023.
[58] Y. Deng, X. Hou, B. Li, J. Wang, and Y. Zhang, “A highly powerful calibration method for robotic smoothing system calibration via using adaptive residual extended Kalman filter,” Robotics and Computer-Integrated Manufacturing, vol. 86, pp. 102660, Jan. 2024.
[59] J. Zhou, H.-N. Nguyen, and H.-J. Kang, “Simultaneous identification of joint compliance and kinematic parameters of industrial robots,” International Journal of Precision Engineering and Manufacturing, vol. 15, pp. 2257–2264, Dec. 2014.
[60] 詹妹臻，利用前饋神經網路與數位雙生學習增強機器手臂的運動學校正演算法，碩士論文，國立台灣大學，機械工程學系，2024
[61] 蕭仁忠，工業機器人定位精度改善之研究，博士論文，國立交通大學，機械工程學系，2020