An Efficient Hierarchical Planner for Autonomous Exploration Based on Ant Colony Path Optimization
H.-Q. Liu, Yankun Wang, Shuo Pei, Guanghui Sun, Weiran Yao · IEEE Sensors Journal · 2025
Current exploration approaches suffer from imbalance between local-coverage-focused exploration and global exploration, as well as increased motion costs due to fragmented or highly angular trajectories. To address these limitations of existing methods, we propose a hierarchical autonomous exploration planner for the intelligent unmanned ground vehicles (UGV) that enhances exploration efficiency while reducing motion costs. At the local planner, a local-coverage-complete strategy is proposed using local frontiers and a localized Rapidly-exploring Random Tree (RRT) to finely explore local regions. At the global planner, a connectivity-aware global route graph (GRG) update and maintenance strategy is proposed based on connected component analysis and historical sampled points and frontiers, thereby improving overall path planning efficiency. To get the optimal exploration path, we compute the information gain estimation of all potential paths, and the metric is extended to frontier evaluation. Besides, to reduce motion costs and improve the path continuity, we propose an Ant Colony Optimization (ACO)-based path optimization method, optimizing both path length and trajectory smoothness to significantly reduce motion costs. Experimental results demonstrate that the proposed exploration method outperforms existing approaches in terms of both exploration efficiency and path quality.