Adaptive Control of Multi-Rotor UAV Using Enhanced Learning Technology

DOI

10.2991/cimns-18.2018.28How to use a DOI?

Keywords

multi-rotor UAV; enhanced learning; PID control; machine learning; self-adaptive control

Abstract

Due to the development and maturation of embedded processor, micro-sensor technology and control theory, multi-rotor UAV is booming. The continuously increasing efficiency and multifunctionality make multi-rotor UAV widely used in military, civil aerial shooting, search and rescue, monitoring and other fields. With the special mechanical structure and dynamic characteristics, it also plays an important role in the scientific research. However, UAV is susceptible to the disturbance of environmental factors during the flight. This study uses reinforcement learning to enhance the stability of flight control of multi-rotor UAV. The reinforcement learning method, also known as reinforcement learning, is one of the learning methods in the field of machine learning and artificial intelligence. It is the most commonly used algorithm in the agent system, which is suitable for the unknown environment. In particular, the agent system will get feedback from changes in the environment, so it can learn from experience and optimize the overall benefit. This research mainly uses the Q-learning algorithm to realize the enhanced learning mechanism, which learns the best response and control corresponding to different UAV attitude, adopts the PID control parameter of reinforcement learning adjustment, and uses it to maintain the stable posture of UAV in the unknown environment.

Copyright

© 2018, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Chin-Hsiung Lee
PY  - 2018/11
DA  - 2018/11
TI  - Adaptive Control of Multi-Rotor UAV Using Enhanced Learning Technology
BT  - Proceedings of the 2018 3rd International Conference on Communications, Information Management and Network Security (CIMNS 2018)
PB  - Atlantis Press
SP  - 124
EP  - 127
SN  - 2352-538X
UR  - https://doi.org/10.2991/cimns-18.2018.28
DO  - 10.2991/cimns-18.2018.28
ID  - Lee2018/11
ER  -