油圧モータ・負荷系へのニューラルネットの適用 : 第3報,ディザーによる定常偏差の改善(機械力学,計測,自動制御)

概要

論文の詳細を見る
The purpose of this study is to develop a precise angular position control system of the hydraulic motor with undesirable dead zone using a Neural Network. In the previous works, we showed that the state feedback Neural Network control has a good effect on improving position control accuracy and robustness, and then, it was confirmed that the application of the PWM signal to the Neural Network position control system attains a very small angular positioning. 0n the other hand, it is well known that one method for eliminating the dead zone characteristics is to superimpose the dither signal into the input current of the servovalve. ln this study, the effective method using a dither signal together with the on-line training of the Neural Network is proposed in order to achieve precise positioning performance. Amplitude and frequency of the dither signal are examined by analyzing a linearized model of the hydraulic motor load system. From the computer simulation and experiments, it is recognized that the dither effect improve the steady performance of Neural Network based position control system.
一般社団法人日本機械学会の論文
2003-08-25