Work

Optimal rapidly-exploring random tree via discrete hierarchical exploration

Year: 2025

Type: article

Abstract: To address the trade-off between timeliness and optimality in complex-environment path planning, this work proposes DHE-RRT*, an enhanced Rapidly-Exploring Random Tree algorithm with discrete hierarchical exploration. The DHE-RRT* algorithm introduces a two-tier recursive exploration framework that discretizes continuous expansion directions into a finite set of headings, thereby facilitating efficient local space coverage. To further enhance sampling efficiency, a forced-neighborhood detection mechanism steers the tree growth toward critical regions shaped by obstacle geometries. Moreover, DHE-RRT* decouples the initial path generation from the subsequent bisection-based topological optimization, effectively reducing computational complexity while maintaining probabilistic completeness. Extensive simulations demonstrate that DHE-RRT* has excellent adaptability and robustness.

Source: Journal of King Saud University - Computer and Information Sciences

DOI: 10.1007/s44443-025-00211-2

Authors: Zeng Zhaoyang, Hu Bowen, Chang Yufang

Institutions: Hubei University of Technology, East China Jiaotong University

Language: English

Topic: Robotic Path Planning Algorithms

Subfield: Computer Vision and Pattern Recognition

Field: Computer Science

Domain: Physical Sciences

Open Access status: gold

Estimated APC paid: $1,990

Funder: National Natural Science Foundation of China