Application of Self-Organizing Map for analyzing robotic arm’s action with Consciousness-Based Architecture module

Authors
Wisanu Jitviriya, Eiji Hayashi
Corresponding Author
Wisanu Jitviriya
Available Online 15 September 2014.
DOI
https://doi.org/10.2991/jrnal.2014.1.2.8
Keywords
Consciousness-Based Architecture (CBA), Human-robot interactions, Self-Organizing Map (SOM)
Abstract
Our research has been focused on developing human-robot interactions, the robots need to perform a high level of intellectual activity and user compatibility. Therefore, we considered the primary structure of a conscious human/animal action, which can be represented by the sequence process, Perception ? Motivation ? Action. We have improved a hierarchical structure model, which defines the relationship between the consciousness field and the behavior module. This model is called Consciousness-Based Architecture (CBA). Furthermore, the robot should select the action itself, we have investigated the application of brain-inspired technology so we introduced a Self-Organizing Map (SOM) neural network that is trained using unsupervised learning to classify its behavior according to the motivation value in the CBA module. In this paper, we attempt to describe the integration of a Self-Organizing Map (SOM) method into a CBA module in order to classify and select autonomous behavior. We confirmed the effectiveness of the proposed system by the experimental results.

Copyright
© 2013, the Authors. Published by ALife Robotics Corp. Ltd.
Open Access
This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)