標題: Low-level autonomous control and tracking of quadrotor using reinforcement learning
作者: Pi, Chen-Huan
Hu, Kai-Chun
Cheng, Stone
Wu, I-Chen
機械工程學系
應用數學系
資訊工程學系
Department of Mechanical Engineering
Department of Applied Mathematics
Department of Computer Science
關鍵字: Reinforcement learning;Policy gradient;Quadrotor
公開日期: 1-Feb-2020
摘要: This paper proposes a low-level quadrotor control algorithm using neural networks with model-free reinforcement learning, then explores the algorithm's capabilities on quadrotor hover and tracking tasks. We provide a new point of view by examining the well-known policy gradient algorithm from reinforcement learning, then relaxing its requirements to improve training efficiency. Without requiring expert demonstrations, the improved algorithm is then applied to train a quadrotor controller with its output directly mapped to four actuators in a simulator, which is a technique used to control any linear or nonlinear system under unknown dynamic parameters and disturbances. We show two experimental tasks both in simulation and real-world quadrotors to verify our method and demonstrate performance: 1) hovering at a fixed position, and 2) tracking along a specific trajectory.
URI: http://dx.doi.org/10.1016/j.conengprac.2019.104222
http://hdl.handle.net/11536/153802
ISSN: 0967-0661
DOI: 10.1016/j.conengprac.2019.104222
期刊: CONTROL ENGINEERING PRACTICE
Volume: 95
起始頁: 0
結束頁: 0
Appears in Collections:Articles