Design, fabrication, and real-time neural network control of a three-degrees-of-freedom nanopositioner

Abstract

A nanometric precision three-degrees-of-freedom positioner is designed and fabricated. Actuation is based on piezoelectric stacks. Capacitive gap sensors with less than 1.0-nm resolution are used for position feedback. In order to design a proper closed-loop controller, the open-loop characteristics of the nanopositioner (static stiffness, hysteresis, drift, frequency response, and the coupling effects) are experimentally investigated. A cerebellar model articulation controller neural network control algorithm was applied in order to provide real-time learning and better tracking capability compared to a standard proportional-integral-derivative control algorithm.