文章基本信息

标题：FUTURE MOTION DECISIONS USING STATE-ACTION PAIR PREDICTIONS
本地全文：下载
作者：Masashi Sugimoto ; Kentarou Kurashige
期刊名称：International Journal of New Computer Architectures and their Applications
印刷版ISSN：2220-9085
出版年度：2015
卷号：5
期号：2
页码：79-93
DOI：10.17781/P001896
出版社：Society of Digital Information and Wireless Communications
摘要：Robots that works in a dynamic environment must possess, the ability to autonomously cope with the changes in the environment. This paper proposes an approach to predict changes in the state and actions of robots. Further, this approach attempts to ap- ply predicted future actions to current actions. This method predicts the robot's state and action for the distant future using the states that the robot adopts repeatedly. Using this method, the actions that the robot will take in the future can be predicted. The method proposed in this paper predicts the state and action of a robot each time it decides to perform an action. In particular, this paper focuses on defin- ing weight coefficients, using the characteristics of the future prediction results. Using this method, the compensatory current action will be obtained. This paper presents the results of our study and discusses methods that allow the robot to quickly determine its most desirable action, using state prediction and optimal control methods.
关键词：Online State Prediction; Prediction using Combina- ; tion of State Space and Action; Online SVR; Con- ; trol using State-action Prediction