Abstract: Recently, magnetic telemanipulation devices have shown a great deal
of promise in such areas as semi-conductor manufacturing, wind tunnels, drug
delivery, and many more. However, these devices are generally associated with
problems caused by payload variation and uncertainties in the parameters of the
system which in turn, have limited the development and application of magnetic
telemanipulation technology to its full capacity. This paper addresses and
deals with these issues by implementation of a precise position control method
for a magnetic telemanipulation system with high level of uncertainties in its
parameters. The levitation system used in this study is primarily designed for
performing remote pick and place operations. The levitated object is a 28
gr microrobot capable of grasping and releasing payloads as heavy as 8
gr. To cope with the uncertainties in the modeling and payload variation, a
model reference adaptive feedback linearization (MRAFL) controller is designed
and its performance compared with an ordinary feedback linearization (FL)
controller. Through experimental results it is shown that the MRAFL controller
enables the microrobot to grasp and transport a payload as heavy as
30% of its own weight without a considerable effect on its positioning
accuracy. In the presence of the payload, the MRAFL controller resulted in a
RMS positioning error of 8 μm} compared with 27.9 μm} of
the FL controller. The approach presented in this work is versatile as it leads
to the modeling and control of a highly nonlinear system through a modular
approach that can be applied to a variety of magnetic levitation and
telemanipulation systems.
