A novel path planning approach for smart cargo ships based on anisotropic fast marching. (30th November 2020)

• Record Type: Journal Article
• Authors: Yan, Xin-ping; Wang, Shu-wu; Ma, Feng; Liu, Yuan-chang; Wang, Jin
• Is Part Of: Expert systems with applications. Volume 159(2020)
• Publication Date: 2020-11-30
• Subjects: Path planning -- Smart ships -- Fast marching method -- Anisotropic fast marching method -- Navigation brain system (NBS); Expert systems (Computer science) -- Periodicals; Systèmes experts (Informatique) -- Périodiques; Electronic journals; 006.33
• DOI: 10.1016/j.eswa.2020.113558
• Languages: English
• ISSNs: 0957-4174

Abstract

Highlights: We analyze the minimum cost problem of path planning. We design an oval shape speed profile for anisotropic fast marching method. We propose a novel potential field model for both static and dynamic obstacles. We propose a novel path planning approach for smart cargo ships. Abstract: Path planning is an essential tool for smart cargo ships that navigate in coastal waters, inland waters or other crowded waters. These ships require expert and intelligent systems to plan safe paths in order to avoid collision with both static and dynamic obstacles. This research proposes a novel path planning approach based on the anisotropic fast marching (FM) method to specifically assist with safe operations in complex marine navigation environments. A repulsive force field is specially produced to describe the safe area distribution surrounding obstacles based on the knowledge of human. In addition, a joint potential field is created to evaluate the travel cost and a gradient descent method is used to search for appropriate paths from the start point to the end point. Meanwhile, the approach can be used to constantly optimize the paths with the help of the expert knowledge in collision avoidance. Particularly, the approach is validated and evaluated through simulations. The obtained results show that it is capable of providing a reasonable and smooth path in a crowded waters. Moreover, the ability of this approach exhibits a significant contribution to the development of expert and intelligent systems in autonomous collision avoidance.