構造的に安定な無駄時間を含む予測制御系の設計

概要

論文の詳細を見る
In order to perform finite settling time control with delay plant, it is necessary for the pole of controlled object to include the pole of reference input. So the pole assignment by state feedback is performed. But this feedback is not continuos but discrete, so hidden oscillation comes out. In order to decrease this oscillation, the samplig time interval must be decreased. The predictive control becomes possible for this delay plant. In the sense of making this transient time least, time optimal control is possible. Not only the past and present time value of the reference input and output but also the future value are necessary. But eternal future values are not necessary. Only several future values are necessary. This is convenient to reaiize this predictive system.
岡山理科大学の論文