近年來無人機被廣泛應用於各個領域，像是農業、物資傳送或者是橋樑檢測等，無人機要在這些領域中能夠順利完成飛行任務避障系統就扮演非常重要的角色，因此本研究的目的為開發一套能夠應用於四旋翼機的避障系統。
在本研究中會建立支援向量機並將其模型和四旋翼機結合，透過支援向量機模型的分類來決定是否要執行避障？一開始會透過感測器蒐集環境數據，並且將這些數據做標準化來提高模型的精準度，要產生一個好的模型除了要有足夠的數據之外，核函數的挑選也很重要，才能夠避免過擬核的狀況產生。藉由感測器的幫忙，當障礙物和無人機的距離位於觸發條件時，會啟動避障機制，在避障的過程中使用者的權限會受到限制，避免干擾的產生。
最後透過飛行測試來驗證此套系統的可行性，由實驗結果中可知無人機能夠在離障礙物3公尺時順利啟動避障機制進行迴避的動作。

Recently, drone is widely used in many application fields such as agriculture, material delivery and bridge inspection. To successfully achieve these goals, avoidance system plays an important role. Therefore, the aim of the thesis is to develop an avoidance system which is applicable to quadcopter.
The thesis will establish the support vector machine(SVM) model and try to combine it with quadcopter. With the help of SVM model, it can decide whether to perform the obstacle collision avoidance according to distance which sensors detect. To get the accurate model, not only sufficient data but also suitable kernel function is also important. At first, sensors collect the data and the will be standardized to increase the accuracy of the model. To select an appropriate kernel function, validation error will be applied to check whether the overfitting will occur or not.
During the process of the obstacle collision avoidance, the priority of the controller is higher than commands from user to reduce unnecessary interference. Finally, the feasibility of avoidance system will be proved by flight test and the results show that the quadcopter can smoothly stop flying about 3 m before obstacle.

參考文獻
[1] E. Dourado, R. Hagemann and A. Thierer, “Operation and Certification of Small Unmanned Aircraft Systems,” Public Interest Comment, Mercatus Center at George Mason University, Arlington, VA, 2015.
[2] A. Shankar, M. Vatsa and P.B. Sujit,“Collision Avoidance for a Low-cost Robot using SVM-based Monocular Vision,” Proceedings of the IEEE International Conference on Robotics and Biomimetics , Dec. 5-10, Bali , Indonesia, pp.277-282, 2014.
[3] Z. Xu, R. Wei, Q. Zhang, K. Zhou and R. He, “Obstacle Avoidance Algorithm for UAVs in Unknown Environment Based on Distributional Perception and Decision Making,” Guidance, Navigation and Control Conference (CGNCC), 2016 IEEE Chinese, pp.1072-1075, 2016.
[4] J. Tian and M. Gao,“Dynamic Collision Avoidance Path Planning for Mobile Robot Based on Multi-sensor Data Fusion by Support Vector Machine,” Proceedings of the IEEE International Conference on Mechatronics and Automation, Aug. 5-8, Harbin, China, 2007.
[5] K.Y. Chee, Z.W. Zhong, “Control, Navigation and Collision Avoidance for an Unmanned Aerial Vehicle,” Sensors and Actuators A:Physical, Vol,190, pp.66-76, 2013.
[6] 吳秉勳, “四旋翼防撞系統設計與驗證研究,” 國立成功大學航空太空工程研究所碩士論文, 2015.
[7] 陳俊維, “應用決策樹演算法於四旋翼防撞系統設計與驗證,” 國立成功大學航空太空工程研究所碩士論文, 2017.
[8] 林軒田, “機器學習技法” , https://www.csie.ntu.edu.tw/~htlin/mooc/
[9] 林軒田, “機器學習基石”, https://www.csie.ntu.edu.tw/~htlin/mooc/
[10] The brief introduction of Time-of-Flight Sensing,
https://www.terabee.com/a-brief-introduction-to-time-of-flight-sensing- part-1-the-basics/
[11] Teraranger Tower Evo Specification,
https://terabee.b-cdn.net/wp-content/uploads/2019/03/TeraRanger-Tower-Evo-Specification-sheet-.pdf
[12] Mission Planner Overview, http://ardupilot.org/planner/
[13] Ardupilot Autopilot Suite, http://ardupilot.com/
[14] C++程式設計,
http://kaiching.org/pydoing/cpp-guide/unit-1-understanding-the-cpp-programming-language.html