Evaluation of robot actions in a continuous state space and planning of action chains using tree structures

2015 Fiscal Year Final Research Report

• Project/Area Number: 24500241
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Perception information processing/Intelligent robotics
• Research Institution: Osaka Prefecture University
• Principal Investigator: Nakashima Tomoharu 大阪府立大学, 工学(系)研究科(研究院), 教授 (20326276)
• Co-Investigator(Kenkyū-buntansha): AKIYAMA Hidehisa 福岡大学, 工学部, 助教 (20533201) ; SEKI Hirosato 大阪大学, 基礎工学研究科, 助教 (10583693)
• Project Period (FY): 2012-04-01 – 2016-03-31
• Keywords: RoboCup / エージェント / 機械学習

Outline of Final Research Achievements

We developed an evaluation method of action chains for robots that are generated in a continuous action space. The proposed method was shown its effectiveness through a series of computational experiments. The best action chains are searched for using a search tree representation. Furthermore, we proposed a strategy representation using kick distribution for optimizing team strategies.

Free Research Field

機械学習