隨著自動化工業已在工業領域中有著不可或缺的地位，而機械手臂是自動化工業中十分重要的一環，其結構由多支連桿所組成，可依照結構不同分成並聯式機械手臂與串聯式機械手臂。機械手臂終端點是主要抓取物品之處，而終端點位置容易受到各連桿的累積誤差，導致終端點定位精度與運行軌跡產生下降，無法達到預期之效果，因此提升終端點的準度是機械手臂相當重要的課題。而為了獲得機械手臂之誤差值，絕大數文獻皆採用絕對精度的機械手臂或雷射追蹤測距儀做為參考，測量機械手臂姿態數據，透過修正運動學之模型參數，提升終端點之靜態定位精度。然而，高精度測量數據儀器的成本非常昂貴，因此本研究中採用自行設計之量測設備，測量機械手臂之運動軌跡數據。
本論文的研究主旨為開發一套系統可以提升機械手臂之軌跡精度，嘗試提升機械手臂的軌跡精確度，使機械手臂運行軌跡可以更接近預期設計之軌跡，藉由實驗的方法來證明其可行性與成效，並探討之未來的發展性與改善目標，其中將分為並聯式機械手臂與串聯式機械手臂來討論。
本論文首先分析並聯式機械手臂與串聯式機械手臂，推導其正向運動學、逆向運動學與動力學，並利用模擬方法對推導結果加以驗證。接著提出一校正與補補償系統之設計方法，採用最小平方法與機械手臂運動學修正機械手臂硬體之參數，接著使用機械手臂動力學與長短期記憶模型建立機械手臂動態誤差模型，預測機械手臂運動過程中真實的軌跡，並且將修正後之角度命令帶入機械手臂控制器中，來提升機械手臂的運動軌跡精準度。並且將修正後之角度命令帶入機械手臂控制器中，控制機械手臂之運轉，再利用3D深度攝影機與慣性感測器，透過影像處理獲得終端點之數據，並利用擴展型卡爾曼方程式，結合兩者感測器之優點，估測出機械手臂終端點實際軌跡，最後對實際數據加以分析，評斷其系統之結果與未來發展性。

The automation industry has an indispensable position in the industrial fields, and a robotic arm is a very important part of the automation industries. Its structure consists of multiple links, and robot arms can be divided into parallel and serial robotic arms according to their structures. The endpoints of the robotic arms are the main place to grab objects, and the positions of the endpoints are easily affected by the cumulative errors by the links. The errors can result in the decreases in the positioning accuracies and moving trajectory of the endpoint, and the expected performances cannot be satisfied. Thus, it is very important for robotic arms to improve the accuracies of the endpoints. In order to obtain the errors of the robot arms, most literatures use an absolute precision robot arm as a reference or a laser rangefinder to obtain the positions and attitudes of robot arms, and then the positions and attitudes are modified by correcting the kinematic model parameters. However, the measurement instruments are expensive. Therefore, this study develops an approach to determine the motion trajectories of robotic arms.
The main purpose of this study is to develop a system that can improve the trajectory accuracies of robot arms, so the trajectories of the robot arms can be closer to the expected trajectories. Besides, the feasibilities of the proposed approach is shown by the experimental results, which are used to discuss the future developments and improvement goals according to parallel and serial manipulators.
This study firstly analyzes a parallel manipulator and a series manipulator, derives their forward kinematics, inverse kinematics and dynamics, and uses the simulation method to verify the derivation results. Secondly, a design method of the calibration and compensation system is proposed, where the least square method and the kinematics of the robot arms are used to correct the parameters of the robot arms. Thirdly, the dynamic error models of the robot arms are established by using the dynamics of the robot arms and the long short-term memory (LSTM) model to predict the motions of the robot arms, which are used to correct the joint angles is added so ass to improve the trajectory accuracies of the robotic arms. Fourthly, the corrected angle commands are delivered to the robot arm controllers, which control the motions of the robot arms. A 3D depth camera and an inertial sensor are used to measure the motions of the endpoint, and then the extended Kalman filter is used to estimate the actual trajectory of the endpoint of the robotic arms by using the data obtained from the two sensors. Finally, the results are analyzed to discuss the future developments of the system.

[1] C-H. Rim, C-M. Rim, J.-G. Kim, G Chen, and J.-S. Pak. “A calibration method of portable coordinate measuring arms by using artifacts,” Mapan-Journal of Metrology Society of India, vol. 34, no. 1, pp. 1–11, 2019.
[2] M. Gaudreault, A. Joubair, and I. Bonev. “Self-calibration of an industrial robot using a novel affordable 3D measuring device,” Sensor, vol. 18, no. 10, pp. 3380, 2018.
[3] Thu T. L. T., Quoc K. D., and Thanh L. P.. “Calibration of industrial robot kinematics based on results of interpolating error by shape function,” Journal of Engineering and Applied Sciences, vol. 15, no. 6, pp. 1451–1461, 2020.
[4] M. Chen, X. Hong, L. Wei, Y. Liu and C. Xu, “Robotic arm calibration and teaching method based on binocular vision,” in 39th Chinese Control Conference, pp.5963–5969, 2020.
[5] G. Li, F. Zhang, Y. Fu, and S. Wang, “Kinematic calibration of serial robot using dual quaternions,” Industrial Robot: the International Joural of Robitics Researach and Application, vol. 46, no. 2, pp. 247–258, 2019.
[6] W. Jiang, X. Li, Z. Luo, L. Yang, B. Guo, and X. Hu, “Calibration of a manipulator with a regularized parameter identification method,” IEEE Access, vol. 10, pp. 90535–90547, 2022.
[7] C. Mao, S. Li, Z. Chen, H. Zu, Z. Wang, and Y. Wang, “A novel algorithm for robust calibration of kinematic manipulators and its experimental validation,” IEEE Access, vol. 7, pp. 90487–90496, 2019.
[8] W. Feng, Y. Tian, M. Ouyang, H. Bian, and Q. Chen, “A positioning error compensation method for multiple degrees of freedom robot arm based on the measured and target position error,” Advances in Mechanical Engineering, vol. 14 no. 5, pp. 1–13, 2022.
[9] M. Maung, M. Latt, and C. New, “DC motor angular position control using PID controller with friction compensation,” International Journal of Scientific and Research Publications, vol. 8, no. 11, pp. 149–155, 2018.
[10] D. Shang, Y. Li, Y. Liu, and S. Cui, “Research on the motion error analysis and compensation strategy of the delta robot,” Mathematics, vol. 7, no. 5 , pp. 411, 2019.
[11] Z. Wang, Z. Chen, C. Mao, and X. Zhang, “An ANN-based precision compensation method for industrial manipulators via optimization of point selection,” Mathematical Problems in Engineering, vol. 2, pp. 2025–2036, 2020.
[12] P. M. Bhatt, R. K. Malhan, P. Rajendran, A. V. Shembekar, and S. K. Gupta, “Trajectory-dependent compensation scheme to reduce manipulator execution errors for manufacturing applications,” in ASME Manufacturing Science and Engineering Conference, June 2021.
[13] C. Yu, Z. Li, and H. Liu, “Research on gravity compensation of robot arm based on model learning,” IEEE/ ASME International Conference on Advanced Intelligent Mechatronics. pp. 635–641, 2019.
[14] H. Su, W. Qi, Y. Hu, J. Sandoval, L. Zhang, Y. Schmirander, G. Chen, A. Aliverti, A. Knoll, G. Ferrigno, and E. D. Momi, ‘‘Towards model-free tool dynamic identification and calibration using multilayer neural network,’’ Sensors, vol. 19, no. 17, p. 3636, 2019.
[15] J. Fugal, J. Bae, and H. Poonawala, “On the impact of gravity compensation on reinforcement learning in goal-reaching tasks for robotic manipulators,” Robotics vol. 10, no. 1, pp. 46, 2021.
[16] D. Nemec, V. Simak, A. Janota, M. Hrubos, and E. Bubenikova. “Precise localization of the mobile wheeled robot using sensor fusion of odometry, visual artificial landmarks and inertial sensors,” Robotics and Autonomous Systems, vol. 112, pp. 168–177, 2019.
[17] S. Mustapha, A. Kassir, K. Hassoun, Z. Dawy, and H. AbiRached, “Estimation of crowd flow and load on pedestrian bridges using machine learning with sensor fusion,” Automation in Construction, vol. 112, pp. 103092.
[18] T. Cvitanic, S. Melkote, and S. Balakirsky. “Improved state estimation of a robot end-effector using laser tracker and inertial sensor fusion,” CIRP Journal of Manufacturing Science and Technology, vol. 38, pp. 51–61, 2022.
[19] N. Shaukat, A. Ali, M. J. Iqbal, M. Moinuddin, and P. Otero, “Multi-sensor fusion for underwater vehicle localization by augmentation of RBF neural network and error-state kalman filter,” Sensors, vol. 21, pp. 1149,. 2021.
[20] D. Sabzevari and A. Chatraei, “INS/GPS sensor fusion based on adaptive fuzzy EKF with sensitivity to disturbances,” IET Radar Sonar Navig, vol. 15, no. 11, pp. 1535–1549, Nov. 2021.
[21] S. Kucuk and Z. Bingul, ‘‘Robot kinematics: Forward and inverse kinematics,’’ in Industrial Robotics: Theory, Modelling and Control. London, U.K.: IntechOpen, 2006, pp. 117–148.
[22] M. J. Thomas, M. L. Joy and A. P. Sudheer, “Kinematic and dynamic analysis of a 3-PRUS spatial parallel manipulator,” Chinese Journal of Mechanical Engineering, vol. 33, no. 1, 2020.
[23] K. E. Clothier and Y. Shang, “A Geometric Approach for Robotic Arm Kinematics with Hardware Design, Electrical Design, and Implementation,” Journal of Robotics, vol. 2010, pp. 1–10, 2010.
[24] L.-W. Tsai, Robot Analysis: The Mechanics of Serial and Parallel Manipulators, John Wiley & Sons: New York, NY, USA, 1999.
[25] M. M. U. Atique, M. R. I. Sarker, and M. A. R. Ahad, “Development of an 8DOF quadruped robot and implementation of inverse kinematics using Denavit-Hartenberg convention,” Heliyon, vol. 4, no. 12, 2018.
[26] P. Corke, “A simple and systematic approach to assigning denavit-hartenberg parameters,” IEEE transactions on robotics, vol. 23, no. 3, pp. 590–594, 2007.
[27] R. Featherstone and D. Orin, “Robot dynamics: equations and algorithms,” in 2000 IEEE International Conference on Robotics and Automation (ICRA). vol. 1, pp. 826–834, 2000.
[28] D. Kostic, B. d. Jager, M. Steinbuch, and R. Hensen, “Modeling and ´ identification for high-performance robot control: An rrr-robotic arm case study,” IEEE Trans on Control System Technology, vol. 12, no. 6, pp. 904–919, 2004.
[29] S. Gale, H. Rahmati, J. T. Gravdahl and H. Martens, “Improvement of a robotic manipulator model based on multivariate residual modeling,” Front. Robot, vol. 4, pp. 28, 2017.
[30] J. L. Aroca Trujillo, R. R. Serrezuela, V. Azhmyakov, and R. S. Zamora, “Kinematic model of the ScorBot 4PC manipulator implemented in Matlab’s Guide,” Contemp. Eng. Sci., vol. 11, no. 4, pp. 183–199, 2018.
[31] L. Perumal, “Euler angles: conversion of arbitrary rotation sequences to specific rotation sequence,” Computer Animation and Virtual Worlds, vol. 25, no. 5, pp. 521–529, 2014.
[32] G. G. Slabaugh, “Computing euler angles from a rotation matrix,” Retrieved August, vol. 6, no. 2000, pp. 39–63, 1999.
[33] D. N. Vila-Rosado and J. A. Dominguez-Lopez, “A matlab toolbox for robotic manipulators,” in 6th Mexican Int. Conf. on Computer Science, pp. 256–263, 2005.
[34] S. Aoyagi, A. Kohama, Y. Nakata, Y. Hayano, and M. Suzuki, “Improvement of robot accuracy by calibrating kinematic model using a laser tracking system-compensation of non-geometric errors using neural networks and selection of optimal measuring points using genetic algorithm,” in International Conference on Intelligent Robots and Systems (IROS), pp. 5660–5665, 2010.
[34] D. W. Marquardt, “An algorithm for least-squares estimation of nonlinear parameters,” J. Soc. Ind. Appl. Math., vol. 11, no. 2, pp. 431–441, 1963.
[35] T. F. Coleman and Y. Y. Li, “An interior trust region approach for nonlinear minimization subject to bounds,” SIAM J. Optim., vol. 6, no. 2, pp. 418–445, 1996.
[36] S. Hochreiter and J. Schmidhuber, “Long short-term memory,” Neural Comput., vol. 9, no. 8, pp. 1735–1780, 1997.
[37] P. J. Werbos, “Backpropagation through time: What it does and how to do it,” Proc. IEEE, vol. 78, no. 10, pp. 1550–1560, 1990.
[38] S. Amari, "Backpropagation and stochastic gradient descent method", Neurocomputing, vol. 5, no. 4-5, pp. 185-196, 1993.
[39] S.-L. Sun and Z.-L. Deng, “Multi-sensor optimal information fusion Kalman filter,” Automatica, vol. 40, no. 6, pp. 1017–1023, 2004.
[40] J. Z. Sasiadek and P. Hartana, “Sensor data fusion using Kalman filter,” in Proc. 3rd Int. Conf. Inf. Fusion, vol. 2, 2000.
[41] J. B. Kuipers, Quaternions and Rotation Sequences: a primer with applications to orbits, aerospace, and virtual reality, Princeton University Press, 1999.
[42] V. A. Chobotov, Spacecraft Attitude Dynamics and Control. Melbourne, FL, USA: Krieger, 1991.
[43] R. M. Haralick, S. R. Sternberg, and X. Zhuang, “Image analysis using mathematical morphology,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 9, no. 4, pp. 532–550, 1987.
[44] H. Yuen, J. Princen, J. Illingworth, and J. Kittler, “Comparative study of hough transform methods for circle finding,” Image and vision computing, vol. 8, no. 1, pp. 71–77, 1990.
[45] H. Yuen, J. Princen, J. Illingworth, and J. Kittler, ‘‘Comparative study of Hough transform methods for circle finding,’’ Image Vis. Comput., vol. 8, no. 1, pp. 71–77, 1990.
[46] M. I. Ribeiro, “Kalman and extended kalman filters: Concept, derivation and properties,” Institute for Systems and Robotics, vol. 43, pp. 46, 2004.
[47] W. T. Wei, Y. F. Tsai, M. Yeh, Y. W. Jan, and Y. W. Wu, “MEMS-based gyro-stellar inertial attitude estimate for NSPO micro-sat program,” Proceedings of IEEE Aerospace Conference, pp. 1–15, 2019.
[48] L. Yao, Y. A. Wu, L. Yao, and Z. Z. Liao, “An integrated IMU and UWB sensor based indoor positioning system,” in Proc. Int. Conf. Indoor Position. Indoor Navig. (IPIN), pp. 1–8, 2017.
[49] P. Axelsson and F. Gustafsson, “Discrete-time solutions to the continuous-time differential Lyapunov equation with applications to Kalman filtering,” IEEE Trans. Autom. Control, vol. 60, no. 3, pp. 632–643, 2015.