1026 ニューラルネットワーク補償器を利用した空気圧シリンダの力制御

概要

論文の詳細を見る
In this paper, I-PD control system with Smith's compensator, the disturbance compensator and the neural network compensator is applied to the force control of the pneumatic cylinder and the effectiveness is discussed. The Smith's method eliminates delay time from the characteristic polynomial, and it is easy to design the control system. However, Smith predictor control system is proposed for the reference-output transfer characteristics and the steady-state error is not zero for the disturbance if the plant is the astatic system, such as the pneumatic cylinder. Also, the Smith's predictor control system contains the modeling error including the friction force. This error causes the big tracking error. The disturbance compensator control disturbances, and the neural network compensator identifies the modeling error and improves the tracking performance in our control system.
一般社団法人日本機械学会の論文
2003-08-05