Swing-up Control of Acrobot System Using Neurocontroller Trained by Genetic Algorithm

カテゴリ: 部門大会

論文No: OS4-3

グループ名: 【C】平成19年電気学会電子・情報・システム部門大会講演論文集

発行日: 2007/09/04

タイトル(英語): Swing-up Control of Acrobot System Using Neurocontroller Trained by Genetic Algorithm

著者名: Eiho Uezato(University of the Ryukyus),Sam Duong Chau (University of the Ryukyus),Hiroshi Kinjo(University of the Ryukyus)

著者名(英語): Eiho Uezato(University of the Ryukyus),Sam Duong Chau (University of the Ryukyus),Hiroshi Kinjo(University of the Ryukyus)

キーワード: acrobot system|neurocontroller|genetic algorithms|nonlinear control|underactuated systems

要約(日本語): For the last several years, the acrobot, a two-degree-of-freedom planar robot with single actuator, have emerged in various literatures as a challenging nonlinear control problem. Various control methods for the acrobot system have been developed including fuzzy approach. This paper presents a new control method which seems to be the first work utilizing neural network and genetic algorithms for the system. Numerical simulation shows that the proposed neurocontroller is able to control the acrobot for swing-up and balance motions.

PDFファイルサイズ: 3,357 Kバイト