Under the Mechanism Design and Control Portion, the
two main components are:
Design
& Analysis of a Star Parallel Mechanism for precise manipulation
In
the past, when products are bulky and precision was not of high importance,
serial link manipulators were often used. With the advancement of
micro-miniaturization and motivation for higher precision, parallel mechanism
becomes popular.
In
this project, the degree of freedom required is 3 as the only the tip is of
concerned. Thus, the Star Parallel mechanism with 3 dof
is proposed. As precision is of high importance, the number of revolute joints
will be minimized as much as possible to prevent backlash. Thus, flexure joints
are used. To be cost effective,
Upon successful implementation of the parallel manipulator, a novel
mechanism will be developed to manipulate and actuate the micro end-effector at the tip of the intraocular shaft.
Successful implementation of this parallel star manipulator will see
advancement in the nano-technology world. For
example, the tip of a nano-manufacturing
equipment can make use of this manipulator to compensate the erroneous motion
due to vibration to achieve a better finish.
Control and Amplification
of Piezoelectric Actuators
Piezoelectric (PZT) actuator is an excellent choice for our
application because of their high bandwidth and high block force. However, a
typical off-the-shelf piezoelectric actuator only has an effective stroke of
10-15 µm. To achieve a workspace of at least 100 µm in all principal
directions, we need an actuation mechanism with 7-10 times of stroke
amplification. Backlashes between mating mechanical parts and space constraint
within a slender handheld instrument are a couple of the major design concerns.
Piezoelectric ceramic material has hysteresis of
about 15% its total stroke, which is a major source of error when using it for
micro-positioning/tracking applications. Hysteresis is a highly non-linear and
complex phenomenon that causes problems in control, especially in
high-frequency (>10 Hz) trajectory tracking application. We propose to
implement a feedforward controller with inverse
rate-dependent hysteresis model. We could model the hysteretic behavior
off-line, and this model possesses a mathematical inverse, we could feedfoward this inverse model to linearize an open-loop
controller to control the piezoelectric actuators. If sub-hundred nanometer of
precision is desired, feedback sensors such as strain-gauges may be used to
close the control loop. The challenge here is to balance the trade-off between
added precision and the loss in system responsiveness.
