呂藝光Leu, Yih-Guang陳勁榮Chen, Jing-Rong2019-09-032020-08-282019-09-032018http://etds.lib.ntnu.edu.tw/cgi-bin/gs32/gsweb.cgi?o=dstdcdr&s=id=%22G060575024H%22.&%22.id.&http://rportal.lib.ntnu.edu.tw:80/handle/20.500.12235/95696本論文自行設計及實現一個兩輪移動平台，其主要架構包含馬達、驅動電路、改變重心的機械結構、避障感測器以及人臉追蹤系統。由於兩輪系統具有時變及高度非線性的特性，因此我們採用比例、積分及微分控制器(Proportional-Integral-Derivative，PID)進行對兩輪移動平台整體的控制，並以機械結構的重心改變使兩輪移動平台前進或後退。 人臉追蹤系統透過平板電腦所收取的影像進行追蹤，並且透過超音波感測器搭配模糊控制作為避障的依據。 針對兩輪移動平台，進行自主平衡、自主移動、人臉追蹤以及避障等實驗，實驗結果顯示本論文所設計的兩輪移動平台確實可行。This thesis designs and implements a self-made two-wheeled mobile platform. The structure of the self-made two-wheeled mobile platform includes the motors and their drive circuits, the mechanism of the change of the center of gravity, obstacle avoidance sensors, and face following module. Because of the time-varying and highly non-linear characteristics of two-wheeled systems, the self-made two-wheeled mobile platform uses the mechanism of the change of the center of gravity to make it move forward or backward, and utilizes the Proportional-Integral-Derivative (PID) to control its balance. The face following modules uses a tablet computer with an image sensor and an image processing to perform face recognition and find a tracking direction. The ultrasonic sensors are combined with the fuzzy control to acquire obstacle avoidance. Some experiments are performed to verify the self-made two-wheeled mobile platform, such as autonomous balance, the movement, face tracking, obstacle avoidance and so forth.PID控制非線性系統人臉辨識PID controlFace recognitionNonlinear system