Related Experiment Videos
Sensor fusion by pseudo information measure: a mobile robot application.
Mohammad Reza Asharif1, Behzad Moshiri, Reza HoseinNezhad
1Department of Electrical and Computer Engineering, Faculty of Engineering, University of Tehran, Iran. asharif@ie.u-ryukyu.ac.jp
ISA Transactions
|August 6, 2002
Summary
This study introduces an enhanced Bayesian fusion method for autonomous mobile robots to create more informative occupancy grid maps. This leads to improved path planning, balancing route length and safety for diverse applications.
Area of Science:
- Robotics
- Artificial Intelligence
- Sensor Fusion
Background:
- Environment perception is critical for autonomous mobile robots.
- Accurate mapping is essential for effective robot navigation and task completion.
Purpose of the Study:
- To develop a novel sensory data integration approach for generating enhanced occupancy grid maps.
- To improve the performance and reliability of environment perception in mobile robotics.
Main Methods:
- An extended Bayesian fusion method for integrating independent sensor information.
- A new neural structure for converting Khepera robot IR sensor data into occupancy probabilities.
- Simulation and experimental validation using cylindrical and Khepera robots.
Main Results:
- The proposed fusion method generates more informative occupancy grid maps compared to existing approaches.
- The resulting maps enable more appropriate and safer path planning.
- A tunable trade-off between route length and safety was achieved through the fusion function.
Conclusions:
- The novel fusion approach significantly enhances environment perception for autonomous robots.
- The method provides a flexible way to tune map characteristics for specific applications.
- This work contributes to more robust and efficient mobile robot navigation systems.