This is the repository for Research and Development Project "Autonomous Exploration for Mobile Robots". The complete system is based on ROS and Hochschule BRS Robocup @Work repository, we implemented some map exploration methods mentioned below:
Methods of Exploration:
- Nearest Frontier [2] Hector_navigation
- Nearest Cluster
- Cost-Utility [1]
- Deep Q-Network
For running these algorithms look at launch file "mas_industrial_robotics/mir_autonomous_exploration/launch/autonomous_exploration_sim.launch" . Also, run rqt_reconfigure and go to hector_exploration_planner settings for changing exploration behavior. You can change exploration behavior from frontier exploration to Cluster Exploration or to cost-utility exploration.
For deep q-network implementation you need to know anaconda. environment file of anaconda is environment.yml . Also, docker files are available.
References
[1] Yan, Z., Fabresse, L., Laval, J., & Bouraqadi, N. (2015). Metrics for performance benchmarking of multi-robot exploration. IEEE International Conference on Intelligent Robots and Systems, 2015–Decem, 3407–3414.
[2] Yamauchi, B. (1998). Frontier-based exploration using multiple robots. Proc. of the Second International Conference on Autonomous Agents, 47–53.
Citation
@misc{daiem2018,
author = {Ali, Daiem Nadir},
title = {Autonommous Exploration for Mobile Robots},
year = {2018},
publisher = {GitHub},
journal = {GitHub repository},
howpublished = {\url{https://github.com/daiemna/Autonomous_Exploration}}
}