簡易檢索 / 詳目顯示

回結果列表
研究生： 林嵩晉
Lin, Song-Jin
論文名稱： 基於閘控自注意力模型與機器人雙手阻抗控制之人機互動系統之研製
Design and Implementation of Human-Robot Interaction System Using Gated Self-Attention Model and Dual-Arm Impedance Control
指導教授： 李祖聖
Li, Tzuu-Hseng S.
學位類別： 碩士
Master
系所名稱： 電機資訊學院 - 電機工程學系
Department of Electrical Engineering
論文出版年： 2021
畢業學年度： 109
語文別： 英文
論文頁數： 93
中文關鍵詞： 雙手服務型機器人 、力量感測器 、阻抗控制 、自注意力模型
外文關鍵詞： Dual-Arm Service Robot, Force Sensor, Impedance Control, Self-Attention Model
相關次數： 點閱：149  下載：0 
分享至:
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報

    • 在人機協作的研究領域中，如何讓機器人即時接收並分析感測器數據，從而做出配合使用者意圖的動作以完成協作任務，為本論文核心議題。本論文將機器人之控制分為上半身之雙手阻抗控制以及下半身之底盤移動狀態預測兩大部分，以完成將大型貨物搬運至指定地點的人機協作任務。安裝在機器人雙手末端之力量感測器在偵測到互動造成的力與力矩之後，該資訊將會交由兩個子系統各自進行分析，其中阻抗控制系統，根據淨外力與淨外力矩輸出末端線性速度與角速度之迭代式，繼而由雙手各自之雅可比矩陣(Jacobian Matrix)映射成馬達之旋轉速度，最後由馬達轉速完成手臂之軌跡規劃。此外，本論文提出一閘控自注意力模型(Gated Self-Attention Model)，採用監督式學習之方式分析力量感測器之時序型資料，利用自注意力模型平行處理的優勢，同時針對時序維度以及資料維度進行自注意力運算，最後由選擇閘控制兩種不同維度特徵的比例，並以凸集合的方式融合兩種特徵，以分類輸出對底盤移動狀態之預測。綜合上述之兩項技術，機器人在配合使用者手部之搬運姿態的同時，亦能隨著使用者的移動意圖改變底盤的方向並移動，此二項技術在本論文所設計之實機實驗當中，亦充分展現其有效性並能夠即時執行人機協作任務。

    In this thesis, a dual-arm impedance control system for carrying big box cooperation and a gated self-attention model for moving intention prediction are proposed to implement a Human-Robot Interaction (HRI) scheme. In the experimental scenario, the user and the service robot must cooperate with each other to transport a big box to the designated position. In the part of upper body, the data will be read from force sensors, which are installed on the end-effectors of manipulators, and the net external forces and torques are transformed into linear and angular velocities of the end-effector by impedance control. Next, end-effector velocities will be mapped to motor velocities by Jacobian matrix, and the trajectories will be planned on the basis of motor velocities. Apart from the upper part, the gated self-attention based moving intention prediction system is proposed in order to make service robot move in accordance with user’s actions. The input data read from force sensors will be preprocessed into time-related serial data, taken as input of model and trained by supervised learning algorithm. Taking the advantage of the ability of parallel processing of self-attention model, the information in time and data dimensions will be extracted simultaneously but independently. The extracted features will be redistributed by selection gate, forming a merged feature in the form of convex combination. After classification, the predicted moving state is obtained and will manipulate the control mobile platform. In the real-time experiments, both systems are demonstrated to be efficient and able to run in real time, and perform the HRI task well and successfully.

    Abstract II Acknowledgement III Contents IV List of Figures VII List of Tables IX Chapter 1 Introduction 1 1.1 Motivation 1 1.2 Related Works 2 1.3 System Overview 4 Chapter 2 Robot Hardware 5 2.1 Introduction 5 2.2 Forward Kinematics of Arms 7 2.3 Mobile Platform 10 2.4 Force Sensor 12 2.5 Pneumatic System 13 Chapter 3 Dual-Arm Cooperation Based on Impedance Control 15 3.1 Introduction 15 3.2 System Overview 16 3.3 Axial Correction 17 3.3.1 Installation Correction 18 3.3.2 Orientation Correction 19 3.3.3 Gravity Compensation 21 3.4 Impedance Control Model 22 3.5 End-Effector Velocity Control 25 3.5.1 Jacobian Matrix in Robotics 26 3.5.2 Control Strategy 28 Chapter 4 Moving Intention Prediction System Using Gated Self-Attention Model 29 4.1 Introduction 29 4.2 Related Network 30 4.2.1 Recurrent Neural Network (RNN) 30 4.2.2 Long Short-Term Memory (LSTM) 32 4.2.3 Gated Recurrent Unit (GRU) 34 4.3 Gated Self-Attention Model 36 4.3.1 System Overview 37 4.3.2 Self-Attention Arithmetic 38 4.3.3 Feed-Forward Layer 45 4.3.4 Residual Summation 45 4.3.5 Gated Convex Summation 47 4.3.6 Multi-Label Classifier 49 4.4 Data Preprocess 52 4.4.1 Data Collection 52 4.4.2 Label Determination 58 4.4.3 Downsampling for Imbalanced Dataset 61 4.5 Training Results 63 4.5.1 Training Settings 63 4.5.2 Structural Comparison 65 4.5.3 Parametric Comparison 68 4.5.4 Attention Matrix Visualization 73 Chapter 5 Real Experiments 75 5.1 Introduction 75 5.2 Dual-Arm Cooperation Verification 76 5.2.1 Error Tracking Results 76 5.2.2 Lifting and Lowering Cooperation 79 5.2.3 Rotation Cooperation 81 5.3 Moving Intention Prediction Verification 83 5.4 Box Transportation Task 85 Chapter 6 Conclusions and Future Work 87 6.1 Conclusions 87 6.2 Future Work 89 References 90

    [1] R. J. Anderson andM. W. Spong, “Hybrid impedance control of robotic manipulators,” IEEE J. Robot. Autom., vol. 4, no. 5, pp. 549–556, 1988.
    [2] S. Schneider and R. H. Cannon, “Object impedance control for cooperative manipulation: Theory and experimental results,” in Proceedings of 1989 International Conference on Robotics and Automation, 1989, pp. 1076–1083.
    [3] R. Kelly, R. Carelli, M. Amestegui, and R. Ortega, “On adaptive impedance control of robot manipulators,” in Proceedings of 1989 International Conference on Robotics and Automation, 1989, pp. 572–577.
    [4] A. A. Goldenberg, “Implementation of force and impedance control in robot manipulators,” in Proceedings of 1988 IEEE International Conference on Robotics and Automation, 1988, pp. 1626–1632.
    [5] L. J. Love andW. J. Book, “Environment estimation for enhanced impedance control,” in Proceedings of 1995 IEEE International Conference on Robotics and Automation, 1995, vol. 2, pp. 1854–1859.
    [6] F. Zeng, J. Xiao, and H. Liu, “Force/torque sensorless compliant control strategy for assembly tasks using a 6-DOF collaborative robot,” IEEE Access, vol. 7, pp. 108795–108805, 2019.
    [7] K. Kosuge and Y. Hirata, “Human-robot interaction,” in Proceedings of 2004 IEEE International Conference on Robotics and Biomimetics, 2004, pp. 8–11.
    [8] R. Ikeura, T. Moriguchi, and K. Mizutani, “Optimal variable impedance control for a robot and its application to lifting an object with a human,” in Proceedings of 11th IEEE International Workshop on Robot and Human Interactive Communication, 2002, pp. 500–505.
    [9] K. Kosuge and N. Kazamura, “Control of a robot handling an object in cooperation with a human,” in Proceedings of 6th IEEE International Workshop on Robot and Human Communication. RO-MAN’97 SENDAI, 1997, pp. 142–147.
    [10] D. J. Agravante, A. Cherubini, A. Bussy, P. Gergondet, and A. Kheddar, “Collaborative human-humanoid carrying using vision and haptic sensing,” in Proceedings of 2014 IEEE International Conference on Robotics and Automation, 2014, pp. 607–612.
    [11] E. Magrini, F. Flacco, and A. DeLuca, “Control of generalized contact motion and force in physical human-robot interaction,” in 2015 IEEE International Conference on Robotics and Automation, 2015, pp. 2298–2304.
    [12] J. Stückler and S. Behnke, “Following human guidance to cooperatively carry a large object,” in Proceedings of 2011 11th IEEE-RAS International Conference on Humanoid Robots, 2011, pp. 218–223.
    [13] V. Duchaine and C. M. Gosselin, “General model of human-robot cooperation using a novel velocity based variable impedance control,” in Second Joint EuroHaptics Conference and Symposium on Haptic Interfaces for Virtual Environment and Teleoperator Systems (WHC’07), 2007, pp. 446–451.
    [14] R. Ikeura and H. Inooka, “Variable impedance control of a robot for cooperation with a human,” in Proceedings of 1995 IEEE International Conference on Robotics and Automation, 1995, vol. 3, pp. 3097–3102.
    [15] L. Peternel, N. Tsagarakis, and A. Ajoudani, “A human–robot co-manipulation approach based on human sensorimotor information,” IEEE Trans. Neural Syst. Rehabil. Eng., vol. 25, no. 7, pp. 811–822, 2017.
    [16] J. Lanini, H. Razavi, J. Urain, and A. Ijspeert, “Human intention detection as a multiclass classification problem: Application in physical human–robot interaction while walking,” IEEE Robot. Autom. Lett., vol. 3, no. 4, pp. 4171–4178, 2018.
    [17] H.-C. Lin, J. Smith, K. K. Babarahmati, N. Dehio, and M.Mistry, “A projected inverse dynamics approach for multi-arm cartesian impedance control,” in Proceedings of 2018 IEEE International Conference on Robotics and Automation, 2018, pp. 5421–5428.
    [18] Y. Ren, Y. Liu, M. Jin, and H. Liu, “Biomimetic object impedance control for dual-arm cooperative 7-DOF manipulators,” Rob. Auton. Syst., vol. 75, pp. 273–287, 2016.
    [19] J. Krüger, G. Schreck, and D. Surdilovic, “Dual arm robot for flexible and cooperative assembly,” CIRP Ann., vol. 60, no. 1, pp. 5–8, 2011.
    [20] H. Sadeghian, F. Ficuciello, L. Villani, and M. Keshmiri, “Global impedance control of dual-arm manipulation for safe interaction,” IFAC Proc. Vol., vol. 45, no. 22, pp. 767–772, 2012.
    [21] T. C. Hsia, “Internal force-based impedance control of dual-arm manipulation of flexible objects,” in Proceedings of 2000 ICRA. Millennium Conference. IEEE International Conference on Robotics and Automation. Symposia Proceedings (Cat. No. 00CH37065), 2000, vol. 4, pp. 3296–3301.
    [22] J. Lee, P. H. Chang, and R. S. Jamisola, “Relative impedance control for dual-arm robots performing asymmetric bimanual tasks,” IEEE Trans. Ind. Electron., vol. 61, no. 7, pp. 3786–3796, 2013.
    [23] F. Caccavale, P. Chiacchio, A. Marino, and L. Villani, “Six-dof impedance control of dual-arm cooperative manipulators,” IEEE/ASME Trans. Mechatronics, vol. 13, no. 5, pp. 576–586, 2008.
    [24] H.S ong, D. Rajan, J. Thiagarajan, and A. Spanias, “Attend and diagnose: Clinical time series analysis using attention models,” in Proceedings of the AAAI Conference on Artificial Intelligence, 2018, vol. 32, no. 1.
    [25] Y.Fang, S.Yang, B.Zhao, and C.Huang, “Cyberbullying detection in social networks using Bi-GRU with self-attention mechanism,” Information, vol. 12, no. 4, p. 171, 2021.
    [26] V. Raheja and J. Tetreault, “Dialogue act classification with context-aware self-attention,” arXiv Prepr. arXiv1904.02594, 2019.
    [27] P. Vinayavekhin, S. Chaudhury, A. Munawar, et al., “Focusing on what is relevant: Time-series learning and understanding using attention,” in Proceedings of 2018 24th International Conference on Pattern Recognition, 2018, pp. 2624–2629.
    [28] Q. Suo, W. Zhong, G. Xun, J. Sun, C. Chen, and A. Zhang, “GLIMA: Global and Local Time Series Imputation with Multi-directional Attention Learning,” in Proceedings of 2020 IEEE International Conference on Big Data, 2020, pp. 798–807.
    [29] Z. Chen, E. Jiaze, X. Zhang, H.S heng, and X. Cheng, “Multi-Task Time Series Forecasting With Shared Attention,” in Proceedings of 2020 International Conference on Data Mining Workshops, 2020, pp. 917–925.
    [30] J. Zheng, A. Xia, L. Shao, T. Wan, and Z. Qin, “Stock volatility prediction based on self-attention networks with social information,” in Proceedings of 2019 IEEE Conference on Computational Intelligence for Financial Engineering & Economics, 2019, pp. 1–7.
    [31] J. Kim, M. El-Khamy, and J. Lee, “T-gsa: Transformer with gaussian-weighted self-attention for speech enhancement,” in Proceedings of ICASSP 2020-2020 IEEE International Conference on Acoustics, Speech and Signal Processing, 2020, pp. 6649–6653.
    [32] J. Pereira and M. Silveira, “Unsupervised anomaly detection in energy time series data using variational recurrent autoencoders with attention,” in Proceedings of 2018 17th IEEE International Conference on Machine Learning and Applications, 2018, pp. 1275–1282.
    [33] N. N. An, N. Q. Thanh, and Y. Liu, “Deep CNNs with self-attention for speaker identification,” IEEE Access, vol. 7, pp. 85327–85337, 2019.
    [34] S. Hochreiter and J. Schmidhuber, “Long short-term memory,” Neural Comput., vol. 9, no. 8, pp. 1735–1780, 1997.
    [35] J. Chung, C. Gulcehre, K. Cho, and Y. Bengio, “Empirical evaluation of gated recurrent neural networks on sequence modeling,” arXiv Prepr. arXiv1412.3555, 2014.
    [36] A. Vaswani, N. Shazeer, N. Parmar, et al., “Attention is all you need,” arXiv Prepr. arXiv1706.03762, 2017.
    [37] K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” in Proceedings of the IEEE conference on computer vision and pattern recognition, 2016, pp. 770–778.
    [38] G. Tsoumakas and I. Katakis, “Multi-label classification: An overview,” Int. J. Data Warehous. Min., vol. 3, no. 3, pp. 1–13, 2007.
    [39] F. Provost, “Machine learning from imbalanced data sets 101,” in Proceedings of the AAAI’2000 Workshop on Imbalanced Data Sets, 2000, vol. 68, no. 2000, pp. 1–3.
    [40] D. P. Kingma and J. Ba, “Adam: a method for stochastic optimization,” Dec. 2014, Accessed: Jun.10, 2021. [Online]. Available: http://arxiv.org/abs/1412.6980.
    [41] J. Duchi, E. Hazan, and Y.S inger, “Adaptive subgradient methods for online learning and stochastic optimization.,” J. Mach. Learn. Res., vol. 12, no. 7, 2011.
    [42] I. Sutskever, J. Martens, G. Dahl, and G. Hinton, “On the importance of initialization and momentum in deep learning,” in International Conference on Machine Learning, 2013, pp. 1139–1147.

    下載圖示
    2026-08-11公開
    QR CODE