Dynamic adaptive intelligent navigation decision making method for multi-object situation in open water. (1st June 2022)

• Record Type: Journal Article
• Authors: He, Yixiong; Liu, Xiao; Zhang, Ke; Mou, Junmin; Liang, Yu; Zhao, Xingya; Wang, Bing; Huang, Liwen
• Is Part Of: Ocean engineering. Volume 253(2022)
• Publication Date: 2022-06-01
• Subjects: Dynamic adaptive -- Intelligent navigation -- Autonomous collision avoidance -- Collision avoidance mechanism -- Traffic environment digitalization; Ocean engineering -- Periodicals; Ocean engineering; Periodicals; 620.4162
• DOI: 10.1016/j.oceaneng.2022.111238
• Languages: English
• ISSNs: 0029-8018

Abstract

Abstract: The intelligent navigation system of autonomous ships aims at making fully autonomous navigation decisions to replace the work of seafarers. Realizing intelligent navigation decision-making in open water is the first stage of a fully autonomous navigation system. Through real-time sampling, rapid update of the own ship's motion parameters, and traffic situation information, an open-water intelligent navigation decision-making method that can dynamically adapt to the residual error in the system and the target ship's random manoeuvre is proposed. A digital traffic environment model comprising different target types is built as the basis of navigation decision-making. An optimal feasible plans of the own ship, which will ensure all targets pass safely, is calculated using the velocity obstacle (VO) theory, and dynamic collision avoidance mechanism to construct an autonomous manoeuvring model. The simulation experiments proved that the combined results of this and previous research could realize autonomous navigation in an open water environment having multiple static and dynamic objects under the premise of satisfying the International Regulations for preventing Collisions at Sea, 1972 (COLREGs) and the good seamanship. This research will provide basic theories and methods to realize the autonomous navigation decision-making of ships in an open water environment with multiple targets. Highlights: A complete framework of intelligent navigation was proposed, which incorporated ship's manoeuvrability, the requirement of CORLEGs, etc. The intelligent navigation method we proposed could adapt to model prediction errors, the target ship's uncoordinated anti-collision, etc. A digital environment that could be used for intelligent navigation decisions was established by object classification. A novel navigation model to keep ships sailing along the planned route was presented, which could produce the target course at any initial condition.