An Efficient Autonomous Exploration Algorithm of Indoor Environment for Mobile Robots Using Topological Map

QI Lizhe,HE Dongdong,CHEN Qian,SUN Yunquan

DOI: https://doi.org/10.13973/j.cnki.robot.220052

2023-01-01

ROBOT

Abstract:An efficient autonomous exploration algorithm TMRRT（topological map based rapidly exploring random tree）is proposed to reduce the times of reaching a known area repeatedly by the mobile robot in the process of autonomous exploration, and thus to improve the efficiency of autonomous exploration. Firstly, the local and global RRTs（rapidly exploring random trees） with variable growth rate are used as the detector to find the map boundary, and cluster the frontier points. Meanwhile, the best detection points are stored as a topological map to avoid reaching the explored area repeatedly by the robot. Finally, the simulations are carried out in different environments and the experiments are carried out in an actual environment. The experimental results show that the average exploration time can be reduced by more than 7.5% and the average path length can be reduced by more than 19.8% by the the proposed exploration algorithm compared with RRT autonomous exploration algorithm, and the average exploration time can be reduced by more than 15.7% and the average path length can be reduced by more than 34.3% compared with FA（frontier-based approach） autonomous exploration algorithm.The results show that the algorithm can effectively improve the efficiency of robot autonomous exploration and is feasible in the actual environment.