自駕車挑戰賽之策略的設計與實現

Design and Implementation of Strategy for AutoRace Challenge

陳暉穎(Hui-Ying Chen)

指導教授 ： 李世安

淡江大學/工學院/電機工程學系人工智慧機器人碩士班/碩士(2022年)

https://doi.org/10.6846/TKU.2022.00573

摘要

本論文提出一個自駕車挑戰賽的策略設計方法，可以有效地改善以下3點問題，(1)機器人在使用硬編碼的跑點策略時，會因為機器人所在起始點不同，導致出界。(2)機器人在判斷燈號時，因為無法正確顏色辨識，造成機器人策略判斷錯誤。(3)機器人完成任務的速度還不夠快，無法在時限內完成比賽。
本文所提出的自主策略方法，能讓使機器人在特定關卡不再只是依循固定的動作，而是能夠運用自身的感測器調整動作完成關卡，進而提升任務成功率。在影像辨識上，本文增加新的顏色判斷演算法，讓機器人的判斷更加精準。並加入機器人速度調整的策略，使機器人在筆直的路線上加快車速，減少完成比賽所需的時間。由實驗結果中，可以驗證本文所提之策略方法確實能改善比賽結果。

關鍵字

自動駕駛 ； 模糊控制 ； 霍夫轉換 ； 影像辨識 ； 避障

並列摘要

This paper proposes a strategy design method for the self-driving challenge, which can effectively improve the following three problems: (1) When the robot uses a fixed running point strategy, the robot will go out of bounds because the starting point of the robot is different. (2) When the robot is judging the traffic light signal, because it cannot identify the correct color, the robot's strategy judgment is wrong. (3) The speed of the robot to complete the task is not fast enough to complete the race within the time limit.
The autonomous strategy method proposed in this paper can make the robot no longer just follow a fixed action in a specific level, but can use its own sensors to adjust the action to complete the level, thereby improving the success rate of the task. In image recognition, this paper adds a new color judgment algorithm to make the robot's judgment more accurate. And add the strategy of robot speed adjustment, so that the robot can pick up speed on a straight route and reduce the time it takes to complete the race. From the experimental results, it can be verified that the strategy method proposed in this paper can indeed improve the race results.

並列關鍵字

autonomous driving ； Fuzzy Control (FC) ； Hough transform ； image recognition ； obstacle avoidance

參考文獻

參考文獻

[1] C. Thorpe, M. H. Hebert, T. Kanade, and S. A. Shafer, “Vision and naviga-tion for the carnegie-mellon navlab,” IEEE Transactions on Pattern Analysisand Machine Intelligence, vol. 10, no. 3, pp. 362–373, 1988.

[2] Navlab1, URL: https://www.ithome.com.tw/news/125897?fbclid=IwAR3_YrpMQnW98CWFiH18knhjBYixx2H_CUiO4Ahwa1tZ3OCuEk7USQsrjWA

[3] AutoRace, URL: https://emanual.robotis.com/docs/en/platform/turtlebot3/autonomous_driving_autorace/

[4] Duckietown, URL: https://www.duckietown.org/