Electromechanical modeling and adaptive feedforward control of a self-sensing scanning fiber endoscope

Abstract

Precise image capture using a mechanical scanning endoscope is framed as a resonant structural-deflection control problem in a novel application. A bipolar piezoelectric selfsensing circuit is introduced to retrofit the piezoelectric tube as a miniature sensor. A data-driven electromechanical modeling approach is presented using system identification and system inversion methods that together represent the first online-adaptive control strategy for the scanning fiber endoscope (SFE). Trajectory tracking experiments show marked improvements in scan accuracy over previous control methods and significantly, the ability of the new method to adapt to changing operating environments.