A multi-layered fast marching method for unmanned surface vehicle path planning in a time-variant maritime environment. (1st January 2017)

• Record Type: Journal Article
• Authors: Song, Rui; Liu, Yuanchang; Bucknall, Richard
• Is Part Of: Ocean engineering. Volume 129(2017)
• Page Start: 301
• Page End: 317
• Publication Date: 2017-01-01
• Subjects: Anisotropic fast marching -- Marine environment -- Path planning -- Surface currents -- Unmanned surface vehicle; Ocean engineering -- Periodicals; Ocean engineering; Periodicals; 620.4162
• DOI: 10.1016/j.oceaneng.2016.11.009
• Languages: English
• ISSNs: 0029-8018

Abstract

Abstract: Concerns regarding the influence of the marine environment, such as surface currents and winds, on autonomous marine vehicles have been raised in recent years. A number of researchers have been working on the development of intelligent path planning algorithms to minimise the negative effects of environmental influences, however most of this work focuses on the platform of autonomous underwater vehicles (AUVs) with very little work on unmanned surface vehicles (USVs). This paper presents a novel multi-layered fast marching (MFM) method developed to generate practical trajectories for USVs when operating in a dynamic environment. This method constructs a synthetic environment framework, which incorporates the information of planning space and surface currents. In terms of the planning space, there are repelling and attracting forces, which are evaluated using an attractive/repulsive vector field construction process . The influence of surface currents is weighted against the obstacles in the planning space using a 4-regime risk strategy . A trajectory is then calculated using the anisotropic fast marching method. The complete algorithm has been tested and validated using simulated surface currents, and the performance of generated trajectories have been evaluated in terms of different optimisation criteria, such as the distance and energy consumption. Highlights: A practical path planning algorithm for unmanned surface vehicle has been developed. The negative influence generated by the environmental factors such as currents has been minimised for path planning. Anisotropic fast marching algorithm is used as the base algorithm to search for an optimal trajectory in a synthetic vector field. A novel obstacle/expectation decision process is proposed to analyse the planning space with the changes of USV's position. A 4-regime risk strategy is proposed to assist with the path planning in a dynamic maritime environment.