2A2-X04 Self-Position Estimate Method by Two Robots Cooperation Using Moveable Landmarks(Swarm Robotics)
Kiyohiko Hattori, Eri Homma, Naoki Tatebe, Toshinori Kagawa · The Proceedings of JSME annual Conference on Robotics and Mechatronics (Robomec) · 2014
Many researchers focus on robot controlling using self-position estimation. One of the ideas of self-positioning estimation is called SLAM, and this approach could improve accuracy of estimation position with a large number of landmarks; on the other hand, it has a problem of computational complexity increasing exponentially. Dealing with this problem, we propose a new method to make maps and measure own position using 2 robots cooperation as moving landmarks each other. This idea make possible to reduce amount of computational complexity dramatically, and it means that in real-time using less power computer and to realize estimate position in high accuracy. In order to validate the method, we make experiment system using real 2 robots in indoor semi ideal environment. As results of experiment, we revealed distance measurement and control error could reduce less than 5%.