Individually-controllable magnetic artificial cilia for microfluidic manipulation tasks

Abstract

This thesis presents the design, modeling, and control of a magnetic artificial cilia system in which the cilia are individually controllable. In nature, cilia exhibit metachronal waves, or a phase difference between adjacent cilia that results in a traveling wave, which may improve pumping performance or efficiency of biological cilia. However, existing magnetic artificial cilia devices typically use actuation by a rotating field generated by Helmholtz coils or by a rotating permanent magnet. These field sources cannot apply a phase shift to the cilia array and therefore cannot generate a metachronal wave. Nevertheless, magnetic actuation remains desirable for cilia devices as it allows for biocompatibility, precise control of sys- tem inputs, and low-cost fabrication of the cilia. In this thesis, a new design for magnetic artificial cilia is presented in which the actuating magnetic field is localized, enabling indi- vidual actuation. However, this design decision leads to challenging research problems in input-pattern identification, nonlinear systems modeling, and control. In addressing these challenges, the contributions of this thesis are to (i) demonstrate that individual control can improve performance in cilia-based devices, (ii) present accurate nonlinear models for pre- dicting the static response, and (iii) develop a machine-learning-based system identification and control strategy for output tracking.