Volume 39, Number 3, June 2021
|Page(s)||477 - 483|
|Published online||09 August 2021|
Path following method for AUV based on Q-Learning and RBF neural network
School of Marine Science and Technology, Northwestern Polytechnical University, Xi’an 710072, China
李乐(1986-), 西北工业大学助理教授, 主要从事水下机器人协同控制研究。e-mail: firstname.lastname@example.org
In the underwater docking process, the oscillation on AUV velocity brings extra challenge on AUV path following. A Q-learning based Sliding Mode Control (SMC) method to increase the path following performances is proposed. Firstly, AUV guidance law is designed to reduce the path following error. Heading and depth sliding mode controllers are designed to track the guidance law. Then, according to AUV velocity, tracking error and the first derivative, the control parameters of SMC are optimized via Q-learning network. RBF neural network is built to accelerate the offline learning rate. Finally, numerical simulations are made to investigate the characteristics of the present method. Comparisons are made between the trained Q-learning based SMC and the traditional SMC. The results show the effectiveness of the present method.
Key words: autonomous underwater vehicle / path following / reinforcement learning / neural network
关键字 : 自主水下航行器 / 路径跟踪 / 强化学习 / RBF神经网络
© 2021 Journal of Northwestern Polytechnical University. All rights reserved.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.