DDPG 알고리즘을 이용한 양팔 매니퓰레이터의 협동작업 경로상의 특이점 회피 경로 계획
Singularity Avoidance Path Planning on Cooperative Task of Dual Manipulator Using DDPG Algorithm
0
조회
0
다운로드
4
인용
0
참고문헌
키워드
초록
When controlling manipulator, degree of freedom is lost in singularity so specific joint velocity does not propagate to the end effector. In addition, control problem occurs because jacobian inverse matrix can not be calculated. To avoid singularity, we apply Deep Deterministic Policy Gradient(DDPG), algorithm of reinforcement learning that rewards behavior according to actions then determines high-reward actions in simulation. DDPG uses off-policy that uses ε-greedy policy for selecting action of current time step and greed policy for the next step. In the simulation, learning is given by negative reward when moving near singulairty, and positive reward when moving away from the singularity and moving to target point. The reward equation consists of distance to target point and singularity, manipulability, and arrival flag. Dual arm manipulators hold long rod at the same time and conduct experiments to avoid singularity by simulated path. In the learning process, if object to be avoided is set as a space rather than point, it is expected that avoidance of obstacles will be possible in future research.
Abstract
When controlling manipulator, degree of freedom is lost in singularity so specific joint velocity does not propagate to the end effector. In addition, control problem occurs because jacobian inverse matrix can not be calculated. To avoid singularity, we apply Deep Deterministic Policy Gradient(DDPG), algorithm of reinforcement learning that rewards behavior according to actions then determines high-reward actions in simulation. DDPG uses off-policy that uses ε-greedy policy for selecting action of current time step and greed policy for the next step. In the simulation, learning is given by negative reward when moving near singulairty, and positive reward when moving away from the singularity and moving to target point. The reward equation consists of distance to target point and singularity, manipulability, and arrival flag. Dual arm manipulators hold long rod at the same time and conduct experiments to avoid singularity by simulated path. In the learning process, if object to be avoided is set as a space rather than point, it is expected that avoidance of obstacles will be possible in future research.
참고문헌 (19)
- 정성엽, "6자유도 매니퓰레이터의 선형 운동에 대한 특이점 회피", 제어.로봇.시스템학회 논문지, 24(11), pp.1025~1032 (2018), 제어·로봇·시스템학회[DOI]
- 박동주, "Artificial Potential Field와 양팔 조작성 지수 기반의 특이점 및 충돌 회피 경로 계획", 전자공학회논문지, 55(5), pp.89~96 (2018), 대한전자공학회
- 김상현, "로봇 메니퓰레이터의 제어를 위한 특이점 회피 알고리즘의 비교 연구", 로봇학회 논문지, 12(1), pp.42~54 (2017), 한국로봇학회
- T. Yoshikawa, "Analysys and control of robot manipulators with redundancy. Robotic Research", The First International Symposium, pp.735~747 (1984)
- HYUNGJU ANDY PARK, "CLOSED-FORM INVERSE KINEMATIC POSITION SOLUTION FOR HUMANOID ROBOTS", International Journal of Humanoid Robotics, 09(03), pp.1250022~ (2012), World Scientific Pub Co Pte Ltd[DOI]
- Pedro Pedrosa Rebouças Filho, "Control of singularity trajectory tracking for robotic manipulator by genetic algorithms", Journal of Computational Science, 30, pp.55~64 (2019), Elsevier BV[DOI]
- Yoshihiko Nakamura, "Inverse Kinematic Solutions With Singularity Robustness for Robot Manipulator Control", Journal of Dynamic Systems, Measurement, and Control, 108(3), pp.163~171 (1986), ASME International[DOI]
- Charles Wampler, "Manipulator Inverse Kinematic Solutions Based on Vector Formulations and Damped Least-Squares Methods", IEEE Transactions on Systems, Man, and Cybernetics, 16(1), pp.93~101 (1986), Institute of Electrical and Electronics Engineers (IEEE)[DOI]
- D. E. Whitney, "The Mathematics of Coordinated Control of Prosthetic Arms and Manipulators", Journal of Dynamic Systems, Measurement, and Control, 94(4), pp.303~309 (1972), ASME International[DOI]
- Wenrui Wang, "An improved artificial potential field method of trajectory planning and obstacle avoidance for redundant manipulators", International Journal of Advanced Robotic Systems, 15(5), pp.172988141879956~ (2018), SAGE Publications[DOI]
- S. R. Buss, "Introduction to inverse kinematics with jacobian transpose, pseudoinverse and damped least squares methods", IEEE Journal of Robotics and Automation, 17, pp.1~19 (2004)
- "3 types of robot singularities and how to avoid them"[URL]
- S. Phaniteja, "A deep reinforcement learning approach for dynamically stable inverse kinematics of humanoid robots", 2017 IEEE International Conference on Robotics and Biomimetics(ROBIO), pp.1818~1823 (2017)
- 강영균, "공 던지기 로봇의 정책 예측 심층 강화학습", 로봇학회 논문지, 15(4), pp.398~403 (2020), 한국로봇학회[DOI]
- 김성운, "백스터 로봇의 시각기반 로봇 팔 조작 딥러닝을 위한 강화학습 알고리즘 구현", 로봇학회 논문지, 14(1), pp.40~49 (2019), 한국로봇학회[DOI]
- 김태원, "현실 세계에서의 로봇 파지 작업을 위한 정책", 가치 심층 강화학습 플랫폼 개발, 2020, pp.15(2), 197~204[DOI]
- Mihai Duguleana, "Obstacle avoidance of redundant manipulators using neural networks based reinforcement learning", Robotics and Computer-Integrated Manufacturing (2011), Elsevier BV[DOI]
- D. L. Pieper, "The kinematics of manipulators under computer control", Ph.D., pp.1968
- L. Rozo, "Learning manipulability ellipsoids for task compatibility in robot manipulation", 2017 IEEE, 2017, 3183~3189