日本機械学会論文集 C編
Online ISSN : 1884-8354
Print ISSN : 0387-5024
ニューラルネットワークを用いた変化球制御型ピッチングマシンの開発
尾田 十八酒井 忍米村 茂河田 憲吾堀川 三郎山本 博之
著者情報
ジャーナル フリー

2005 年 71 巻 702 号 p. 555-560

詳細
抄録

The most common commercial pitching machines are the “two rollers” type and also the “arm” type. These machines tend to have certain limitations. In particular it is very difficult to simultaneously change both ball speed and direction. Also some types of pitches, such as the curve or sinking ball, are not easily achieved. In this study, the hardware and software design of a new “intelligent” pitching machine which is able to pitch repeatably with selectable speed, direction and ball rotation, is presented. The machine had three rollers and the motion of each was independantly controlled by a heirarchical neural network. This network accepted ball speed, direction and rotation as inputs and produced detailed motion control of the three rollers as output.

著者関連情報
© 社団法人日本機械学会
前の記事 次の記事
feedback
Top