Abstract

The Rapidly-exploring Random Tree (RRT) based method has been widely used in the field of autonomous exploration of mobile robots. During autonomous exploration, the robot relies on the RRT tree growth to obtain exploration frontiers. However, due to the trap space problem, such as narrow corridors and mazes, the RRT tree cannot grow to these regions in a limited time, resulting in the slow extraction of frontiers. Moreover, limited to the randomness of the growth of the RRT tree, the frontiers extracted from the RRT treetop have a lot of redundancy. In this paper, we propose a Generalized Voronoi Diagram (GVD) based exploration method to guide robots for efficient exploration. First, a lightweight feature extraction is proposed to extract the GVD information and represent the topological structure of the environment. Second, the GVD features are utilized to quickly generate heuristic frontiers. In order to reduce the redundancy of frontiers, a GVD frontiers fusion and extraction algorithm is proposed. Finally, the GVD nodes are fused and we obtain a feature node set. By using the collision-check module, we get the feature matrix. The path cost is quickly calculated by using the feature matrix. The experimental results show that by comparing with the RRT-Exploration algorithm, our method has better performance in quickly extracting exploration frontiers and reducing backtracking.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call