抄録
Since modeling based on physical analysis of an ultra-sonic motor(USM) is difficult, conventional PID or IMC-PID control often has been used. However, it is difficult to accomplish satisfactory control performance using a conventional PID or IMC-PID control because an USM contains dynamic variations and non-linear elements.
In this paper, we propose IMC-PID control using neural networks(NN) for an USM. In the proposed method, one parameter of IMC-PID control is adjusted using NN at each sampling-time. Introduction of NN makes it possible to compensate dynamic variations and non-linearity of an USM. The effectiveness of the proposed method is confirmed by experiment using existent USM.
