𝔖 Bobbio Scriptorium
✦   LIBER   ✦

Collision avoidance in multi-robot systems based on multi-layered reinforcement learning

✍ Scribed by Yoshikazu Arai; Teruo Fujii; Hajime Asama; Hayato Kaetsu; Isao Endo


Publisher
Elsevier Science
Year
1999
Tongue
English
Weight
812 KB
Volume
29
Category
Article
ISSN
0921-8890

No coin nor oath required. For personal study only.

✦ Synopsis


It is important for a robot to acquire adaptive behaviors for avoiding surrounding robots and obstacles in complicated environments. Although the introduction of a learning scheme is expected to be one of the solutions for this purpose, a large size of memory and a large calculation cost are required to handle useful information such as motions of robots. In this paper, we introduce the multi-layered reinforcement learning method. By dividing a learning curriculum into multiple layers, the number of expected situations can be reduced. It is shown that real robots can adaptively avoid collision with each other and to obstacles in a complicated situation.