논문리뷰 & 로드맵 정하기 DCULab 이웅원. 1. Real-Time Motion Planning for Agile ~

Presentation on theme: "논문리뷰 & 로드맵 정하기 DCULab 이웅원. 1. Real-Time Motion Planning for Agile ~"— Presentation transcript:

1 논문리뷰 & 로드맵 정하기 2016.09.06 DCULab 이웅원

2 1. Real-Time Motion Planning for Agile ~

3 1. Real-Time Motion Planning for Agile ~ 2002 developing an algorithm that enables the robot to move from its original location to a new location Cell decomposition methods roadmap methods artificial potential field methods

4 1. Real-Time Motion Planning for Agile ~ Cell decomposition methods http://slideplayer.com/slide/3294196/

5 1. Real-Time Motion Planning for Agile ~ roadmap methods

6 1. Real-Time Motion Planning for Agile ~ artificial potential field methods

7 1. Real-Time Motion Planning for Agile ~ Example of tree expansion procedure

8 2. Autonomous Exploration in Unknown Urban Environments for Unmanned Aerial Vehicles - 2005 autonomous exploration method for unmanned aerial vehicles in unknown urban environment. combining model predictive control (MPC) with a local obstacle map builder. An onboard laser scanner is used to build the online map

9 2. Autonomous Exploration in Unknown Urban Environments for Unmanned Aerial Vehicles - 2005 Dynamic programming  cost function “distance from obstacle”

10 2. Autonomous Exploration in Unknown Urban Environments for Unmanned Aerial Vehicles - 2005

11 3. 3D Obstacle Avoidance in Adversarial Environments for Unmanned Aerial Vehicles pan/tilt mounted laser rangefinder is explored as a means of identifying and characterizing potential obstacles in three dimensions (3D)

12 3. 3D Obstacle Avoidance in Adversarial Environments for Unmanned Aerial Vehicles pan/tilt mounted laser rangefinder is explored as a means of identifying and characterizing potential obstacles in three dimensions (3D)

13 4. First Results in Detecting and Avoiding Frontal Obstacles from a Monocular Camera for Micro Unmanned Aerial Vehicles Camera-based methods Various Cues from image (1) monocular (2) disparity (3) optical flow 카메라 하나로 depth 정보추출 정확히는 depth 가 아닌 충돌까지의 시간

14 4. First Results in Detecting and Avoiding Frontal Obstacles from a Monocular Camera for Micro Unmanned Aerial Vehicles

15 1. 실내호버링 https://youtu.be/CccoyyX-xtE PX4Flow  http://dev.px4.io/flow_lidar_setup.htmlhttp://dev.px4.io/flow_lidar_setup.html 250mm Frame 에 pixhawk 를 장착하고 PX4Flow 센서를 달고 실내에서 호버링 해볼 것 * Simulation 은 따로 필요없음

16 2. 장애물 감지  IR sensor https://www.youtube.com/watch?v=WUs2IV1kdUw https://github.com/Cyrilanthony777/Obstacle_Avoidance_Multir otorhttps://github.com/Cyrilanthony777/Obstacle_Avoidance_Multir otor http://www.flitetest.com/articles/quadcopter-obstacle-avoidance- system-using-arduinohttp://www.flitetest.com/articles/quadcopter-obstacle-avoidance- system-using-arduino IR sensor + Ultrasonic sensor  detect obstacle and back off Gazebo Simulation 으로 먼저 장애물 감지

17 3. 장애물 피하기  path planning 싱가폴 대학의 장애물을 피해서 가기 https://www.youtube.com/watch?v=M05zQFlWjDg Path planning algorithm 예시 https://www.youtube.com/watch?v=jn9qZJR6WJ0 https://www.youtube.com/watch?v=63bS1Wv9Jcw

18 9월9월실내 호버링 + 장애물 감지, 데모데이준비 Gazebo 에서 장애물감지 (IR sensor, Ultrasonic Sensor,..) 10 월장애물 피하는 path planning Path planning 공부, gazebo 에서 장애물 회피하는 path planning 해보기 11 월 Vision 공부, vision-based control Monocular, stereo-vision, 비전으로 호버링하기 비전으로 장애물 감지 12 월 Vision 으로 장애물 회피하는 path planning Indoor, outdoor  고정익, 회전익