Autonomous Exploration Algorithms for Unmanned Aerial Vehicles in Unknown Environments
Fan Yang · 2025
Autonomous exploration in GNSS-free indoor scenarios is a hot issue in UAV research. Existing autonomous exploration algorithms for UAVs have been able to run smoothly on small on-board computers and accomplish tasks while meeting the requirements of exploration efficiency and accuracy. However, most of the current algorithms fail to solve the problem of duplicate routes caused by the fact that “corner clusters” near obstacles and “closed-loop clusters” wrapped by known regions are not prioritized for exploration. In this paper, we summarize the common characteristics of these clusters and propose a new FIS structure, which adds the data of the number of voxels in the cluster, the plane size of the voxels, and the ratio of the number of voxels in the cluster located in the effective distance field on the basis of the original FIS structure, and gives the corresponding judgment method. Based on the FUEL Planner autonomous exploration algorithm, an improved algorithm to drive rotorcraft for autonomous exploration, based on the first layer of the algorithm to solve the global path with the traveler algorithm, we add a sequence of clusters with greater priority to place the “edge clusters” and “closed-loop clusters”, and prioritize the access to this sequence. “The sequence of clusters in this sequence is accessed first, and then the clusters solved by the original FUEL algorithm using the traveler algorithm are accessed. The algorithm is simulated and verified against the original FUEL Planner algorithm.