Fast path planning for underwater robots by combining goal-biased Gaussian sampling with focused optimal search. (October 2021)

• Record Type: Journal Article
• Authors: Shen, Jie; Fu, Xiao; Wang, Huibin; Shen, Shaohong
• Is Part Of: Computers & electrical engineering. Volume 95(2021)
• Publication Date: 2021-10
• Subjects: Autonomous path planning -- Robots navigation -- Rapidly-exploring Random Tree Star -- Goal-biased Gaussian distribution sampling -- Focused optimal searching; Computer engineering -- Periodicals; Electrical engineering -- Periodicals; Electrical engineering -- Data processing -- Periodicals; Ordinateurs -- Conception et construction -- Périodiques; Électrotechnique -- Périodiques; Électrotechnique -- Informatique -- Périodiques; Computer engineering; Electrical engineering; Electrical engineering -- Data processing; Periodicals; Electronic journals; 621.302854
• DOI: 10.1016/j.compeleceng.2021.107412
• Languages: English
• ISSNs: 0045-7906

Abstract

Abstract: Autonomous path planning plays an important role in the navigation of intelligent underwater robots. Path planning is a nondeterministic polynomial hard issue in classical path planning models. This problem can be solved using various sample-based strategies. However, the effectiveness of these sample-based strategies is significantly lower in underwater environments, owing to the special undulating terrain and obstacles that are sparser compared to those in the ground. In this study, a more efficient underwater path planning method is proposed for underwater robot navigation. The method employs a goal-biased Gaussian sampling algorithm to select searching nodes optimally, and a focused optimal search algorithm is proposed to accelerate the path optimization process. Combining these two algorithms results in high-efficiency and fast autonomous underwater path planning. Experimental results demonstrate that our method can generate a shorter path and is more efficient than a rapidly exploring random tree star in underwater robot navigation. Highlights: Path planning is a nondeterministic polynomial hard issue in classical path planning models. Underwater path planning methods should be more efficient because many calculations performed using previous methods are unnecessary. This method can regularize the sampling process and accelerate the convergence speed of path optimization for robot navigation. A goal-biased Gaussian sampling strategy is used with variable standard deviations, which can quickly initialize the path to the goal. The focused optimal search algorithm was applied to minimize the initial path. Graphical abstract: Image, graphical abstract