Autonomous exploration of a mobile vechicle using frontier_based algorithm for the exploration and RRT/RRT*/A*/Hybrid A* for the move_base. Mobile robot in an unknown environment has to explore the environment autonomously and create a map for it, for which ros package gmapping and a lidar named sick_tim are used. Then frontier_based algorithm would apply a new destination for robot to keep exploring and move_base package generates a viable trajectory from current positon to the destination. Several options are applied when selecting the trajectory generation algorithms including RRT, RRT*, A* and Hybrid A*. In Gazebo, the localization of the mobile vehicle can be obtained from Gazebo while the encoders are used in the real test. In addition to the encoders, ros package laser_scan_matcher is used for better localization.

• ubuntu 18.04
• ROS is installed
• gmapping

download the code directly, and directory autonomous_exploration is considered as the workspace.
git clone https://github.com/HAOLI-TUKL/autonomous_exploration.git

1. enter the workspace cd autonomous_exploration

2. compile the workspace catkin_make

3. source before run source devel/setup.bash

4. call the launch file to run all the nodes roslaunch mycar launch_all.launch

• my_frontier_exploration package generates a destination(frontier) for mobile vehicle in order to approach the boundary between the explorated and unexplorated area. When multiple destinations(frontiers) can be selected, a closest one in the DFS tree would be chosen.
• my_move_base package generates a trajectory from the current position to the destination generated by my_frontier_exploration package. The trajectory can be generated based on RRT/RRT* / A* / Hybrid A*, which can be determined in the launch_all.launch. Then PID controller would be used to have the vehicle follow the trajectory.
• my_car package and my_controller mainly deals with the vehicle model and driver for the simulation in Gazebo.

HAO LI [email protected]
JIEMING CHEN

Yamauchi, Brian. “A frontier-based approach for autonomous exploration.” Proceedings 1997 IEEE International Symposium on Computational Intelligence in Robotics and Automation CIRA'97. 'Towards New Computational Principles for Robotics and Automation' (1997): 146-151.