A Skeleton-Based Topological Planner for Exploration in Complex Unknown Environments

Haochen Niu,Xingwu Ji,Lantao Zhang,Fei Wen,Rendong Ying,Peilin Liu
2024-12-18
Abstract:The capability of autonomous exploration in complex, unknown environments is important in many robotic applications. While recent research on autonomous exploration have achieved much progress, there are still limitations, e.g., existing methods relying on greedy heuristics or optimal path planning are often hindered by repetitive paths and high computational demands. To address such limitations, we propose a novel exploration framework that utilizes the global topology information of observed environment to improve exploration efficiency while reducing computational overhead. Specifically, global information is utilized based on a skeletal topological graph representation of the environment geometry. We first propose an incremental skeleton extraction method based on wavefront propagation, based on which we then design an approach to generate a lightweight topological graph that can effectively capture the environment's structural characteristics. Building upon this, we introduce a finite state machine that leverages the topological structure to efficiently plan coverage paths, which can substantially mitigate the back-and-forth maneuvers (BFMs) problem. Experimental results demonstrate the superiority of our method in comparison with state-of-the-art methods. The source code will be made publicly available at: \url{<a class="link-external link-https" href="https://github.com/Haochen-Niu/STGPlanner" rel="external noopener nofollow">this https URL</a>}.
Robotics
What problem does this paper attempt to address?
This paper attempts to solve the problems of inefficient path planning and high computational cost encountered during autonomous exploration in complex and unknown environments. Specifically, existing methods rely on greedy heuristic algorithms or optimal path planning, which are prone to result in repeated paths (BFMs) and high computational requirements, thus affecting the exploration efficiency. To address these issues, the authors propose a new exploration framework, STGPlanner, which utilizes the global topological information of the environment to improve exploration efficiency and reduce computational overhead. The following are the main contributions of the paper: 1. **New autonomous exploration framework**: By using the skeleton topological graph (STG) representation of the environment geometry, the problem of repeated paths is effectively solved. 2. **Incremental skeleton extraction method**: Based on the wave - front propagation method, it can efficiently construct a concise environmental topological structure. 3. **STG - based exploration strategy**: A finite - state machine (FSM) is designed to fully utilize STG information to make efficient exploration decisions and reduce computational cost. 4. **Extensive experimental evaluation**: Demonstrates the effectiveness of this method in various environments. ### Specific problems and solutions #### 1. Inefficient path planning - **Problem**: Existing methods rely on local information for path planning and are prone to fall into inefficient repeated paths (BFMs), especially in complex environments. - **Solution**: By introducing global topological information, the robot is guided to prioritize the completion of the exploration of the current branch and avoid unnecessary backtracking behavior. #### 2. High computational cost - **Problem**: Frequent re - planning of the optimal path (such as TSP) leads to an excessive computational burden. - **Solution**: Use a lightweight STG to represent the environment and make efficient path decisions through FSM, reducing the need for complex computations. ### Method overview 1. **Incremental skeleton extraction**: Extract the skeleton from the known environment to form the basis of STG. 2. **STG update**: Expand and simplify STG according to the current exploration branch and the updated skeleton. 3. **Exploration strategy FSM**: Based on STG information, achieve efficient exploration decisions through state transitions. Through these methods, STGPlanner not only improves exploration efficiency but also significantly reduces computational cost. Experimental results show that this method performs well in multiple environments and has high practical value.