IIP情報・知能・精密機器部門講演会講演論文集
Online ISSN : 2424-3140
セッションID: G-1-1
会議情報
G-1-1 ニューラルネットワークによる音響的フィードバックシステムを用いたジョイスティック操作力の定値制御(メカニカルシステムとその知能化(1),口頭発表)
小倉 治佐藤 太一五十嵐 洋
著者情報
会議録・要旨集 フリー

詳細
抄録
The experimental device was set up for measuring human's joystick operating force.The joystick was fixed on a table and test subjects sat on a chair and operated the joystick to the sound.Test subjects listen to white noise output from a speaker.We have established an acoustic control system that feedbacks operating force for realizing target value of joystick operating force.Firstly,we have tried to use the neural network to shorten the time for calibration of human.The duration of this calibration is 9 seconds.Compared with the time duration of calibration in the PI control reported in our previous study,the time for calibration experiment considerably decreased.We conducted a constant-value control experiment using acoustic control system with neural network right after calibration experiment.The neural network methods we applied have better characteristics in the constant-value control experiment than PI control method.
著者関連情報
© 2013 一般社団法人 日本機械学会
前の記事 次の記事
feedback
Top