A Bidirectional APF-RRT* Path Planning Algorithm with Multi-Strategy Adaptive Expansion

Guoliang Cheng, Jiuxiang Yan, Qian Li, Yongguo Zhao, Malgat Mazakov, Kaliyeva Kulyash

2025

Closed access Metadata (CSL-JSON)
DOI 10.1109/mlprae67267.2025.11291064
Data confidence 70 medium
How this score is calculated

It is built from checkable properties of the record itself: whether the article has a DOI, how many independent sources confirmed its data, whether the required fields are filled in, and how recent the latest confirmation is. Disagreement between sources and a missing identifier lower it. A value of 80 or above means the record was published without further checking; 60 to 79 means it is published at reduced confidence, with a manual check advisable. A record verified by an editor by hand is never pushed below 90 by a recalculation. This scores our data about the article, not the article itself.

Abstract

For intelligent systems such as mobile robots and unmanned aerial vehicles, path planning is a core component, and path quality and time efficiency are important indicators for evaluating path planning. Currently, Rapidly-exploring Random Tree (RRT)-based algorithms used for mobile robot path planning still face issues such as high randomness in the search space, slow convergence, and resulting curvy paths. To address these problems, this paper proposes an improved Bidirectional Artificial Potential Field-RRT* (IBi-APF-RRT*) path planning algorithm. In node expansion, the method first combines random tree expansion with a goal-biasing strategy to eliminate the randomness of the tree and rapidly direct the search toward the goal. Additionally, based on bidirectional exploration, an optimized artificial potential field (APF) method is employed to guide the tree expansion toward the goal region while avoiding obstacles. Experimental results show that the proposed IBi-APF-RRT* algorithm outperforms several popular RRT* variants in terms of both path length and runtime. Compared to RRT*, APF-RRT*, Bi-RRT*, and Bias-RRT*, the average path length is reduced by 15.22%, 5.46%, 17.17%, and 11.25%, respectively, and the average runtime is reduced by 95.39%, 94.20%, 52.47%, and 51.01%, respectively, demonstrating the efficiency and adaptability of the improved algorithm.