Recurrent neural network control for LCC-resonant ultrasonic motor drive

A newly designed driving circuit for the traveling wave-type ultrasonic mot or (USM), which consists of a push-pull DC-DC power converter and a two-pha se voltage source inverter using one inductance and two capacitances (LCC) resonant technique, is presented in this study. Moreover, because the dynam ic characteristics of the USM are difficult to obtain and the motor paramet ers are time varying, a recurrent neural network (RNN) controller is propos ed to control the USM drive system. In the proposed controller, the dynamic backpropagation algorithm is adopted to train the RNN on-line using the pr oposed delta adaptation law. Furthermore, to guarantee the convergence of t racking error, analytical methods based on a discrete-type Lyapunov functio n are proposed to determine the varied learning rates for the training of t he RNN. Finally, the effectiveness of the RNN-controlled USM drive system i s demonstrated by some experimental results.