臺灣博碩士論文加值系統

近年來具跟隨功能之搬運車逐漸流行，市面上出現許多相關產品 。這些相關產品多以影像識別技術實現跟隨功能，在人群較擁擠的環境容易發生跟隨錯誤標的之情況。為改善此問題，本研究採用 超寬頻(Ultra-Wideband, UWB)定位技術實現具跟隨功能之搬運車。
本文實作之跟隨車使用無風扇電腦做為執行平台，並採用機器人作業系統(Robot Operating System, ROS)來和底層之馬達控制板溝通。本研究根據麥克納姆輪運動模型之特性開發驅動程式，藉此驅動由麥克納姆輪所承載之車體，使跟隨車可以全向移動，達到靈活移動之效果。跟隨車車體三處各配置一個UWB錨點，並於使用者身上配置一個UWB標籤。UWB標籤接收到UWB錨點之訊號 ，可以藉由該訊號之媒體存取控制(Media Access Control, MAC)位址確認跟隨標的是否正確，因此跟隨之精準度較不受人群影響。接著，本研究採用三邊定位 演算法將錨點與標籤之距離資料換算成使用者相對於跟隨車之位置，再驅動車體往使用者位置移動，以實現跟隨功能。
本跟隨車亦有實現ROS機器人的重要基本功能：建立地圖、導航、避障。這些功能需要精準的里程資訊(Odometry)與速度控制，為此，本研究採用將霍爾編碼器與六軸慣性測量單元所測得之資料進行融合以提供里程資訊，並利用PID(Proportional Integral Derivative)控制程式調整速度使車體移動較為平順。上述功能亦需搭配準確的周遭環境資訊，因此本跟隨車之車體前端有裝設一光學雷達來收集環境資料，之後透過座標轉換架構(Transform Frame)，將環境資料由雷達坐標系轉換為車體中心點坐標系。最後，本研究亦建置一款行動應用程式，方便多名使用者同時觀看跟隨車狀態和位置。本研究所實現之跟隨車可提供跟隨、建立地圖、導航、及避障等功能，並可利用手持裝置即時觀看跟隨車所在位置。期待未來可將其應用到生活中，分擔繁重之工作。

In recent years, following vehicles become more and more popular. There are also many related products available on the market. Most of them are realized using image recognition technology which tends to miss targets in crowded environments. To solve the problem, this thesis realized a following vehicle using an Ultra-Wideband (UWB) localization.
A software named Robot Operating System (ROS) is executed on a fanless minicomputer in the proposed vehicle. The ROS communicates with a motor control board for motion control. A driver program is developed based on a kinematic model of a four Mecanum wheeled vehicle so that the proposed vehicle can move in any direction. Three positions of the vehicle are equipped with one UWB anchor. The person to be followed carry an UWB tag whose Media Access Control (MAC) address can be exploited to identify the target. It is the key reason that the proposed solution works well in crowded environments. Then, a triangulation algorithm is designed to estimate the position of the target. Finally, the four Mecanum wheels are driven to move the vehicle toward the target.
The proposed vehicle further implemented some fundamental functionalities of ROS such as Simultaneous localization and mapping (SLAM), navigation, and obstacle avoidance. Those functionalities require not only highly precise odometry but also smooth motion control without compromising accuracy. Therefore, a data fusion approach is designed to provide the odometry estimation through fusing data from four Hall encoders and one 6-axis inertial measurement unit. Besides, a Proportional Integral Derivative (PID) control program is adopted to adjust the velocity for achieving smooth robot motion. The above functionalities also require precise position information of obstacles around the vehicle. Hence, a LiDAR is mounted on the front of the vehicle to collect the information which are, then, transformed from the LiDAR coordinate system to the vehicle base coordinate system. Finally, the research also built a mobile application for multi-user to check the status of the vehicle concurrently.

摘要...i
Abstract...ii
誌謝...iv
目錄...v
表目錄...vii
圖目錄...viii
第1章 緒論...1
1.1 研究背景與動機...1
1.2 問題描述...1
1.3 研究回顧...2
1.3.1 機器人作業系統...2
1.3.2 Simultaneous Localization and Mapping...2
1.3.3 Navigation stack...3
1.3.4 室內定位技術...3
1.3.5 具跟隨技術之相關產品...4
1.4 論文架構說明...4
第2章 跟隨車設計...5
2.1 設計概述...5
2.2 機體結構...5
2.3 底盤設計...7
2.4 馬達結構...9
第3章 運動分析...11
3.1 底盤運動學模型...11
第4章 硬體系統架構...14
4.1 系統概述...14
4.2 中央控制系統與主要軟體...15
4.2.1 光學雷達Lidar...15
4.2.2 UWB室內定位套件...16
4.3 微控制單元系統...17
4.4 電力系統...17
第5章 實驗方法與設計...21
5.1 方法概述...21
5.2 車體控制...21
5.2.1 PID參數控制...22
5.3 SLAM...24
5.4 Navigation...26
5.5 跟隨機制與架構...29
5.5.1 基於光學雷達的障礙物探測...29
5.5.2 UWB定位機制...29
5.5.3 跟隨...31
5.6 APP設計與架構...33
5.6.1 APP分頁說明...33
第6章 實驗結果...35
6.1 基礎功能測試...35
6.2 SLAM測試...46
6.3 Navigation測試...50
6.4 跟隨測試...52
6.5 APP成果..57
第7章 結論與未來展望...60
參考文獻...61
Extended Abstract...63

[1] Boston Dynamics Available from: https://www.bostondynamics.com/spot
[2] ROS org Available from: https://www.ros.org/
[3] Giorgio Grisetti, Cyrill Stachniss, and Wolfram Burgard, 2007 ,“Improved Techniques for Grid Mapping With Rao-Blackwellized Particle Filters”, IEEE Transactions on Robotics,.23(1):p. 34-36
[4] Smith, R. C.、Self, M.、Cheeseman, P.,1986, “Estimating Uncertain Spatial Relationships in Robotics”, Appears in Proceedings of the Second Conference on Uncertainty in Artificial Intelligence,435-461
[5] Andrew J. Davison; Ian D. Reid; Nicholas D. Molton; Olivier Stasse, 2007 , “MonoSLAM: Real-Time Single Camera SLAM”, IEEE Transactions on Pattern Analysis and Machine Intelligence:p.1052-1067
[6] Raul Mur-Artal, J. M. M. Montiel, Juan D. Tardos,2015.” ORB-SLAM: a Versatile and Accurate Monocular SLAM System”, IEEE Transactions on Robotics,p.17. 13 figures
[7] Stefan Kohlbrecher. Oskar von Stryk. Johannes Meyer. Uwe Klingauf, 2011, “A Flexible and Scalable SLAM System with Full 3D Motion Estimation”, IEEE International Symposium
[8] Wolfgang Hess, Damon Kohler, Holger Rapp, Daniel Andor, 2016, “Real-Time Loop Closure in 2D LIDAR SLAM”, 2016 IEEE International Conference on Robotics and Automation
[9] Dieter Fox Damon Kohler, Wolfram Burgard, 1997, “The Dynamic Window Approach to Collision Avoidance”, IEEE Robotics & Automation Magazine 4(1),p.23-33
[10] D. V. Lu, D. Hershberger and W. D. Smart, 2014,“Layered costmaps for context-sensitive navigation”, 2014 IEEE/RSJ International Conference on Intelligent Robots and Systems, Chicago,2014
[11] 陳宗逸,眾規格原理拼高下，無線室內定位應用大爆發,2018 年版通訊產業關鍵報告, https://www.2cm.com.tw/2cm/zh-tw/market/89B526D1A3C34CE9BD6ACBD31B8D95BF
[12]Passion Mobility U200 Available from: http://www.passionmobility.com/Products/product_info/ukart/7
[13]東元電機F-200系列: https://www.teco.com.tw/assets/pdf/Intelligent_following_car_tw.pdf
[14] 郡野實業有限公司 (建東機器) 智慧型自動跟隨搬運採收車: https://www.kblin.com.tw/product-detail-2447511.html
[15] cowarobot Available from:https://cowarobotrover.com/
[16] 九號平衡車Plus :https://www.mi.com/scooterplus
[17] Hamid Taheri, Bing Qiao, Nurallah Ghaeminezhad, 2015, “Kinematic Model of a Four Mecanum Wheeled Mobile Robot”, 2015 International Journal of Computer Applications
[18] ROS navigatiom/tutoriais/robotsetup/odom Available from: http://wiki.ros.org/navigation/Tutorials/RobotSetup/Odom
[19] Linorobot Available from: https://github.com/linorobot/linorobot
[20] Attitude and heading reference system Available from: https://en.wikipedia.org/wiki/Attitude_and_heading_reference_system
[21] Introduction to PID Controllers Available from: http://georgegillard.com/documents/2-introduction-to-pid-controllers
[22] Yoonseok Pyo, Hancheol Cho, Leon Jung, Darby Lim, 2017,ROS Robot Programming,11, ROBOTIS
[23] amcl-ROS wiki Available from: http://wiki.ros.org/amcl
[24] F. Dellaert. D. Fox. W. Burgard. S. Thrun,2002, “Monte Carlo localization for mobile robots”, IEEE International Conference on Robotics and Automation (ICRA)
[25] Kullback–Leibler divergence Available from: https://en.wikipedia.org/wiki/Kullback%E2%80%93Leibler_divergence
[26] ROS indoor localization Available from: http://wiki.ros.org/indoor_localization
[27] ROS message filters Available from: http://wiki.ros.org/message_filters
[28] Android_ROS Wiki Available from: http://wiki.ros.org/android
[29] rosjava/Android_core Available from: https://github.com/rosjava/android_core
[30] org.ros.android.MasterChooser Class Reference Available from: http://docs.ros.org/en/hydro/api/android_core/html/classorg_1_1ros_1_1android_1_1MasterChooser.html
[31] org.ros.android.view.visualization.VisualizationView Class Reference Available from: http://docs.ros.org/en/hydro/api/android_core/html/classorg_1_1ros_1_1android_1_1view_1_1visualization_1_1VisualizationView.html
[32]右手定則 Available from: https://en.wikipedia.org/wiki/Right-hand_rule
[33] Goran Huski´c, Sebastian Buck, Luis Azareel Ibarg¨uen Gonz´alez and Andreas Zell, 2017,” Outdoor Person Following at Higher Speeds using a Skid-Steered Mobile Robot”, IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Canada, September 24–28
[34] D. Nistér, O. Naroditsky, J. Bergen,2006,” Visual odometry for ground vehicle applications”, IEEE Computer Society Conference on Computer Vision and Pattern Recognition , vol. 1, pp. I–652 (2004)