Name: 四旋翼飛行器驚人嘅運動能力
Uploaded: 2013-06-11T19:44:05Z
Duration: 15 min 54 s
Description: 喺 TEDGlobal 機器人實驗室，Raffaello D'Andrea 展示佢嘅四旋翼飛行器。飛行器有自主學習能力嘅算法，可以好似運動員咁樣思考，同埋解決物理運動嘅問題。佢精彩示範咗無人機點樣接住物件、平衡，同埋做決定。留意動力控制四旋翼飛行器呢一幕，因為呢部飛行器勁到令你好想要返部。

Nonlinear quadrocopter attitude control

A platform for aerial robotics research and demonstration: The Flying Machine Arena

A flying inverted pendulum

Stability and control of a quadrocopter despite the complete loss of one, two, or three propellers

Adaptive fast open-loop maneuvers for quadrocopters

A computationally efﬁcient algorithm for state-to-state quadrocopter trajectory generation and feasibility veriﬁcation

Cooperative quadrocopter ball throwing and catching

Admittance control for physical human-quadrocopter interaction

Raffaello D'Andrea: 四旋翼飛行器驚人嘅運動能力

四旋翼飛行器驚人嘅運動能力

喺 TEDGlobal 機器人實驗室，Raffaello D'Andrea 展示佢嘅四旋翼飛行器。飛行器有自主學習能力嘅算法，可以好似運動員咁樣思考，同埋解決物理運動嘅問題。佢精彩示範咗無人機點樣接住物件、平衡，同埋做決定。留意動力控制四旋翼飛行器呢一幕，因為呢部飛行器勁到令你好想要返部。

Quadcopters can now perform feats once reserved for athletes, like balancing poles, catching balls, and even flying with damaged propellers. Their agility comes from algorithms that use physics and control theory to learn and adapt, allowing these machines to make split-second decisions and recover from unexpected setbacks. In one demonstration, a quadcopter adjusted its moves every 20 milliseconds to hit a flying ball, showing how machines can master athletic skills through rapid calculation and practice. By combining machine athleticism with human gestures, new forms of collaboration and interaction are emerging.