臺灣博碩士論文加值系統

近年來自動駕駛的技術發展得愈來愈成熟，藉由整合機器人定位、底盤控制、影像辨識、光學雷達雷射等許多的技術整合成一個系統，來讓自駕車能夠安全地行駛在道路上，但是目前大多數的自駕車研究需要昂貴的硬體設備為研究門檻，例如一台電動車、大量的傳感器以及人工智慧超級電腦等設備，使得大多數機器人研究者只對室內機器人進行研究。

因應送餐等服務業在一般道路的移動需求，但考量行駛於公用道路車道上的法律限制，因此本研究以行人的移動方式為基礎，開發可行走於人行道或道路邊的全自主外送機器人。為了要達到此一目的。透過光學鏡頭與光學雷達等感測器，辨識並分析可用於行人行走的人行道、道路邊、斑馬線等區域，確保機器人不會影響路上車流、不會撞擊行人並且能夠自主導航到達目的。

因此本研究開發了行走在道路邊之全自主外送機器人，本研究使用ROS機器人系統來開發系統，系統中分為兩大方法即時定位與地圖建構(Simultaneous Localization And Mapping, SLAM)和導航(Navigation)，為了處理光學雷達的大量資料，在建立地圖時的機器人能夠精確定位，以規劃出正確航路來建立出精確地圖。而導航方法為航點跟隨功能和利用路面邊線辨識實現道路邊矯正功能。

機器人的執行平台 使用 Nvidia Jetson AGX Xavier作為系統核心，底盤架構採用2WD架構使機器人移動，使用光學鏡頭辨識路面邊線，光學雷達來收集環境資料來建立地圖和導航時動態偵測障礙物，IMU幫助機器人定位。機器人進行導航時跟隨航點行駛在虛擬道路上，使用3D Li-DAR偵測與避開障礙物和使用光學鏡頭辨識路面邊線來盡可能行走在道路邊，來使機器人能夠安全地到達目的地。此一研究成果可實現機器人在道路邊的自主導航與避障，並有潛力可應用於送餐和送貨等服務。

In recent years, the development of autonomous driving technology is becoming more and more mature. There are many technologies such as robot positioning, chassis control, image recognition, LiDAR laser, etc included. It is worth noting that the most self-driving requires higher level hardware equipments, such as electric cars, multiple sensors, and artificial intelligence supercomputers, etc., That makes the most robotic researchers staying the researches of indoor robots.
In order to reach the general road movement requirements of the service such as food delivery. However, considering aboult the legal restrictions Therefore, we developed an unmanned food delivery robot that could move on the sidewalk or road side .The optical lens and 3D LiDAR sensors are used in this study top, to developed an autonomous food delivery robot. This research is beased on the ROS robot system. In order to process a large amount of 3D LiDAR data, the robot can be accurately positioned when building the map. This research proposed a possibity of autonomous navigation and obstacle avoidance of robots on the side of the road with the potential to provid the delivery services.

摘要..........i
Abstract..........ii
誌謝..........iii
目錄..........iv
表目錄..........vii
圖目錄..........viii
第一章 緒論..........1
1.1 研究背景..........1
1.2 研究動機與目的..........1
1.3 研究回顧..........2
1.3.1 即時定位與地圖建構(SLAM)..........2
1.3.2 路徑規劃..........3
1.4 問題描述..........3
1.5 論文架構說明..........4
第二章 戶外機器人分析..........5
2.1 戶外機器人分析與探討一..........5
2.2 戶外機器人分析與探討二..........8
2.3 光學雷達..........11
2.3.1 雷射測距..........11
2.4 航點跟隨..........12
2.5 線跟隨..........12
2.6 霍夫變換..........13
2.7 光源影響分析..........14
2.7.1 白平衡分析..........14
2.8 道路邊狀況分析..........15
2.8.1 行人可行走道路分析..........15
2.8.2 道路最外側狀況分析..........16
2.9 交通事件防範對策分析..........17
2.9.1 開車門不慎交通事件分析..........17
2.9.2 閃爍警示燈分析..........18
2.10 戶外機器人分析..........18
第三章 硬體設計與架構..........19
3.1 硬體設計概述..........19
3.2 機械結構設計..........19
3.3 底盤輪子材質選擇..........21
3.4 動力機構設計..........24
3.5 照明設備選擇..........25
3.5.1 機器人前燈..........25
3.5.2 機器人後燈..........26
3.6 硬體系統架構概述..........27
3.7 上層系統OS..........28
3.7.1 機器人系統核心..........28
3.8 下層執行單元..........29
3.8.1 環境偵測系統..........29
3.8.2 輸入輸出介面設備..........31
3.8.3 影像辨識系統..........31
3.8.4 第一代攝影機..........31
3.8.5 第二代攝影機..........32
3.8.6 時間定時系統..........33
3.8.7 底盤驅動單元..........35
3.8.8 電力系統..........36
第四章 軟體系統架構..........38
4.1 SLAM即時定位與地圖構建..........38
4.2 Navigation導航與避障..........39
4.2.1 Navigation系統..........39
4.2.2 成本地圖..........40
4.2.3 路徑規劃..........41
4.3 移動機器人定位系統..........43
4.3.1 移動機器人定位..........43
4.3.2 擴展卡爾曼濾波器..........44
4.3.3 自適應蒙特卡羅演算法..........44
4.3.4 TF轉換..........45
4.4 點雲轉換為雷射掃描..........46
4.5 路面邊線辨識..........48
4.6 道路邊矯正演算法..........49
4.7 手機APP IP 攝像頭傳送至ROS系統..........50
4.7.1 獲得手機影像..........50
4.7.2 手機影像傳送至ROS 系統..........51
4.8 感測器融合..........52
第五章 實驗方法設計與實現..........55
5.1 戶外資料收集與電腦規格..........55
5.1.1 戶外收集之資料..........55
5.1.2 顯卡驅動與顯卡型號..........57
5.1.3 運行演算法之電腦規格..........57
5.2 雷射掃描測試與更改..........58
5.2.1 雷射掃描偵測物體測試..........58
5.2.2 雷射掃描更改..........61
5.3 Cartographer的優化..........62
5.3.1 配置文件的參數調整..........63
5.4 IMU的校正..........65
5.5 SLAM Cartographer ..........66
5.5.1 有無IMU校正的地圖建構..........67
5.5.2 戶外地圖驗證..........69
5.6 戶外路面邊線辨識..........71
5.6.1 影像處理之霍夫曼概率..........71
5.6.2 戶外不同光源之路面邊線辨識測試..........75
5.7 時間同步下的擴展卡爾曼濾波器..........78
5.8 機器人導航與避障..........79
5.8.1 光學雷達感知範圍..........79
5.8.2 航點設置..........80
5.8.3 航點跟隨測試..........82
5.8.4 航點跟隨暫停與重啟動功能..........84
5.8.5 避開障礙物測試..........85
第六章 實驗結果..........90
6.1 戶外航點跟隨測試..........90
6.1.1 尋找道路邊的航點..........90
6.1.2 安全路徑測試的場地與地圖..........92
6.1.3 安全路徑測試航點設置..........93
6.1.4 安全路徑測試..........95
6.1.5 避障測試的場地與地圖..........97
6.1.6 避障測試航點設置..........100
6.1.7 避障測試航點距離設置..........102
6.1.8 避障測試成本地圖設置..........104
6.1.9 避障測試..........105
6.1.10 在道路邊的行走測試..........111
6.2 戶外道路邊矯正測試..........116
6.2.1 道路邊矯正測試設置..........116
6.2.2 道路邊矯正測試..........116
第七章 結論與未來展望..........119
7.1 結論..........119
7.2 未來展望..........119
參考文獻..........121
Extended Abstract ..........124

1.Global Autonomous Cars Data. Available from: https://www.bnext.com.tw/article/61531/autonomous-cars-ev.
2.Global Autonomous Cars Current Situation. Available from: https://www.viatech.com/tw/2020/11/self-driver-tw/.
3.Global Autonomous Cars MPI（Miles Per Intervention）Rank. Available from: https://kknews.cc/zh-tw/car/4b93b5g.html.
4.Grisetti, G., C. Stachniss, and W. Burgard, Improved techniques for grid mapping with rao-blackwellized particle filters. IEEE transactions on Robotics, 2007. 23(1): p. 34..
5.Wolfgang Hess, Damon Kohler, Holger Rapp,and Daniel Andor,Real-Time Loop Closure in 2D LIDAR SLAM.
6. 無人載具室內建圖技術 Available from:https://collegeplus.itri.org.tw/2020/03/27/%E7%84%A1%E4%BA%BA%E8%BC%89%E5%85%B7%E5%AE%A4%E5%85%A7%E5%BB%BA%E5%9C%96%E6%8A%80%E8%A1%93/.
7.Hart, P.E., N.J. Nilsson, and B. Raphael, A formal basis for the heuristic determination of minimum cost paths. IEEE transactions on Systems Science and Cybernetics, 1968.4(2): p. 100-107.
8.Padgett, C. and K. Kreutz-Delgado, A grid algorithm for autonomous star identification.IEEE Transactions on Aerospace and Electronic systems, 1997. 33(1): p. 202-213.
9.Koenig, S. and M. Likhachev, Fast replanning for navigation in unknown terrain. IEEE Transactions on Robotics, 2005. 21(3): p. 354-363.
10.Fox, D., W. Burgard, and S. Thrun, The dynamic window approach to collision avoidance. IEEE Robotics & Automation Magazine, 1997. 4(1): p. 23-33.
11.ROS Melodic Package. Available from: http://wiki.ros.org/melodic.
12.鍾毅宏., 智慧型機器人室內戶外定位導航與路徑規劃.2010.
13.林宜臻., 基於深度學習之戶外導航機器人.2018.
14.雷射雷達LIDAR工作原理 Available from: https://twgreatdaily.com/bnU0qWwBvvf6VcSZenMH.html.
15.VLP-16 User Manual. Available from: https://velodynelidar.com/wp-content/uploads/2019/12/63-9243-Rev-E-VLP-16-User-Manual.pdf.
16.全球定位系统 Available from: https://zh.wikipedia.org/zh-tw/%E5%85%A8%E7%90%83%E5%AE%9A%E4%BD%8D%E7%B3%BB%E7%BB%9F.
17.GPS封裝格式與$GPRMC解析 Available from: http://blog.ittraining.com.tw/2016/02/gpsgprmc.html.
18.ROS Melodic Package Available from: http://wiki.ros.org/follow_waypoints.
19.ROS Melodic Tutorial Available from: https://u.cs.biu.ac.il/~yehoshr1/89-685/.
20.opencv中的影象矩（空間矩，中心矩，歸一化中心矩，Hu矩） Available from: https://codertw.com/%E7%A8%8B%E5%BC%8F%E8%AA%9E%E8%A8%80/576722/.
21.霍夫變換 Available from: https://zh.wikipedia.org/wiki/%E9%9C%8D%E5%A4%AB%E5%8F%98%E6%8D%A2.
22.相機基本知識 #6：白平衡 Available from: https://snapshot.canon-asia.com/taiwan/article/zh/camera-basics-6-white-balance?gclid=CjwKCAjww-CGBhALEiwAQzWxOvm1KmqYkgE7h-3--diyWpUrQs3tkyP2q-m53OW7TpDqMzSSw0uEfRoCcUgQAvD_BwE.
23.與「惡」的距離?安全開關車門五步驟!! Available from: https://www.ntpc.gov.tw/ch/home.jsp?id=28&dataserno=201909030001.
24.守護夜間行車！公路爆閃警 南投要設600座 Available from: https://news.ltn.com.tw/news/life/breakingnews/2393491
25.輪子種類有哪些？８種常見輪子材質 Available from: https://www.hocaster.com.tw/blog-detail/%E8%BC%AA%E5%AD%90%E6%9D%90%E8%B3%AA%E4%BB%8B%E7%B4%B9-i.9.
26.Definition of at the top of the hour. Available from:https://www.merriam-webster.com/dictionary/at%20the%20top%20of%20the%20hour.

27.ROS Melodic Package. Available from:https://github.com/cartographer-project/cartographer_ros.
28.Algorithm walkthrough for tuning. Available from:https://google-cartographer-ros.readthedocs.io/en/latest/algo_walkthrough.html.
29.Navigation Package. Available from: http://wiki.ros.org/navigation.
30.Lu, D.V. Navigation Illumination - Shedding Light on the ROS Navstack. Available from: https://roscon.ros.org/2014/wp-content/uploads/2014/07/ROSCON2014_DLu.pdf.
31.Monte Carlo localization. Available from: https://en.wikipedia.org/wiki/Monte_Carlo_localization.
32. YAML. Available from https://zh.wikipedia.org/wiki/YAML.

33.Thrun, S., W. Burgard, and D. Fox, Probabilistic robotics. 2005: MIT press.
34. ROS Melodic Package. Available from http://wiki.ros.org/robot_pose_ekf/Tutorials/AddingGpsSensor.
35. ROS Melodic Package. Available from http://wiki.ros.org/amcl.
36.pointcloud_to_laserscan Package. Available from: http://wiki.ros.org/pointcloud_to_laserscan.
37.OpenCV Package. Available from:https://github.com/opencv/opencv.
38.IP摄像头 Available from: https://play.google.com/store/apps/details?id=com.pas.webcam&hl=zh_TW&gl=US.
39.ROS Melodic Package. Available from: https://github.com/AndresFlorez-Git/IP_Camera_pkg.
40.Velodyne-Lidar-Sync Setup from: https://manual.dewesoft.com/x/setupmodule/devices/velodyne-lidar/velodyne-lidar-setup/velodyne-lidar-sync.
41.message_filters Available from:http://wiki.ros.org/message_filters.
42.Rosbag Available from: http://wiki.ros.org/rosbag.
43.Topics Available from: http://wiki.ros.org/rosbag.
44.X.Org Server Available from: https://zh.wikipedia.org/zh-tw/X.Org_Server.
45. google 地圖 Available from: https://www.google.com/maps/place/%E5%9C%8B%E7%AB%8B%E8%99%8E%E5%B0%BE%E7%A7%91%E6%8A%80%E5%A4%A7%E5%AD%B8/@23.7026158,120.4281881,17z/data=!3m1!4b1!4m5!3m4!1s0x346eba072ee22f11:0x558aed05907207d7!8m2!3d23.7026129!4d120.4294895
46.Tuning methodology cartographer. Available from: https://github.com/dpkoch/imu_calib.
47.imu_calib. Available from: https://google-cartographer-ros.readthedocs.io/en/latest/tuning.html.
48.四元數Available from: https://zh.wikipedia.org/wiki/%E5%9B%9B%E5%85%83%E6%95%B8.
49.Writing a tf listener (C++). Available from http://wiki.ros.org/tf/Tutorials/Writing%20a%20tf%20listener%20%28C%2B%2B%29.
50.GIMP. Available from https://zh.wikipedia.org/wiki/GIMP.
51.道路交通安全規則Available from https://law.moj.gov.tw/LawClass/LawSingleRela.aspx?media=print&PCODE=K0040013&FLNO=80&ty=L.