On-line trained neural network controller for ultrasonic motor servo drive

An on-line trained neural network model-following controller is proposed to control the rotor position of an ultrasonic motor (USM) servo drive in this paper. First, a new design driving circuit for the USM, which is a two-phase chopper-inverter combination, is introduced. Next, the proposed neural network controller is designed and implemented. Accurate tracking response can be obtained by random initialization of the neural network weights due to the powerful on-line learning capability, moreover, the influences of parameter variations and external disturbances of the USM servo drive can be effectively reduced by the proposed neural network controller.