Design and control of piezoelectric-driven handheld manipulator for tremor compensation in microsurgery

• Main Author: Tan, U Xuan
• Other Authors: Ang Wei Tech
• Format: Theses and Dissertations
• Language: English
• Published: 2010
• Subjects: DRNTU::Engineering::Mechanical engineering::Robots; DRNTU::Engineering::Mechanical engineering::Surgical assistive technology
• Online Access: https://hdl.handle.net/10356/40882
• Institution: Nanyang Technological University

The problem of correcting imprecision due to physiological hand tremor while performing micromanipulation tasks such as microsurgery has received considerable attention. This results in the development of a hand-held instrument, Micron, that senses its own motion using an inertial measurement unit; distinguishes intended and undesired motion, and actively attenuates the undesired motion by an opposite but equal deflection of the tip through a piezoelectric-driven manipulator. This dissertation emphasizes an improved version of the aforementioned instrument with regards to the design and control of the piezoelectric-driven mechanism. The hysteretic behavior of piezoelectric actuators makes control challenging. Hence, Prandtl-Ishlinskii (PI) hysteresis model is used to model this hysteretic effect. However, there is a limitation in the PI model. The inverse does not exist when the slope of the hysteretic curve is not positive definite. The author proposes over- coming this problem by mapping the hysteresis through a transformation onto a singularity-free domain, where the inversion can be obtained. The proposed solution brings along a number advantages to the instrument: 1) a stable controller as the limitation of the inverse model is removed; 2) better accuracy as smaller threshold values can be used; and 3) accelerometers will no longer pick up impacts caused by the jerks at the turning point due to the singularity problem. This solution is a general method and can be extended to model any hysteretic materials or systems. The hysteretic behavior is dependent on the individual actuator and also environmental conditions like temperature. Hence, the actuator has to be modeled regularly. To overcome this, the author proposes closing the feedback loop to obtain an adaptive rate-dependent feedforward controller for piezoelectric actuators. The tracking performance of the adaptive controller is performed on a piezoelectric actuator and found to produce a smaller tracking error. In addition, since it is adaptive, experiments are no longer needed to be conducted to obtain the actuator’s model and precious time is saved. This dissertation also presents a new piezoelectric-driven 3-DOF flexure-based par- allel mechanism. To reduce the cost and also to allow complexity in the design to reduce the size, a rapid prototyping machine (Objet) is used to build the mechanism portion as a single piece.