原文传递
Constructing an Autonomous Decision Approach in a Real AIS Environment Using Deep Reinforcement Learning
| 题名： | Constructing an Autonomous Decision Approach in a Real AIS Environment Using Deep Reinforcement Learning |
| 正文语种： | 英文 |
| 作者： | Ruolan ZHANG, Masao FURUSHO |
| 作者单位： | 神戸大学大学院海事科学研究科, 神戸大学大学院海事科学研究科 |
| 关键词： | AIS, decision, re in forcemen t learning, data processing, frame motion, coiiision avoidance |
| 摘要： | Shipbome AIS devices is a global automatic identification system (AIS) that can transmit large amounts of information such as cross obstacles, weather resistance, ship-to-shore, ship-to-ship information exchange, and close-range target displays. In this paper, we present a real historical AIS environment, combined with a rule-based decision and a neural-based decision, and construct an autonomous collision avoidance decision system. Rule based decision-making has several applications in the field of adaptive systems, expert systems, and decision support systems, including automatic ship navigation. However, with the growing amount of data and the expansion of some complex application scenarios, including automation and unmanned situation, it has become progressively challenging to satisfy autonomous decision-making development requirements. The main part of this paper, we examine the performance of decision neural networks under different input types of AIS data. We present a discussion on tlie various training requirements and different training precisions for text data and frame data. Moreover, we include instructive guidance for processing radar, camera, VHF and other training data types. |
| 出版年： | 2020 |
| 期刊名称： | 日本航海学会誌 |
| 卷： | 1月 |
| 期： | 第211号 |
| 页码： | 37-38 |
相关文献
- Online longitudinal trajectory planning for connected and autonomous vehicles in mixed traffic flow with deep reinforcement learning approach
- Multiagent Soft Actor-Critic for Traffic Light Timing
- Random Prior Network for Autonomous Driving Decision-Making Based on Reinforcement Learning
- Where is supply chain software headed?
- Hierarchical Driving Strategy for Connected and Autonomous Vehicles Making a Protected Left Turn at Signalized Intersections
检索历史
应用推荐
