Towards fused vision and force robust feedback control of nanorobotic-based manipulation and grasping

摘要

This paper presents the design and development of a fused vision force feedback robust controller for a nanomanipulator used in nanofiber grasping and nano-fabric production applications. The RRP (revolute revolute prismatic) nanomanipulator considered here utilizes two rotational motors with 0.1 urad resolution and one linear Nanomotor with 0.25 nm resolution. The nanomanipulator is capable of achieving well-behaved kinematic and backlash-free characteristics in addition to atomic scale precision to guarantee accurate manipulation at the nanoscale. A mathematical model of the nanomanipulator is formulated, along with direct and inverse kinematics and dynamic equations of the system. Unlike typical macroscale manipulator models and controllers, the controller development is not trivial due to nanoscale movement and forces, coupled with unmodeled dynamics, nonlinear structural dynamics and mainly the lack of position and velocity feedbacks in this particular nanomanipulator. Utilizing pure vision feedback to acquire nanomanipulator's tip position is shown to have limited capabilities in achieving the desired control accuracy. To remedy this, a controlled fusion of visual servoing and force feedbacks for nanomanipulator positioning is proposed and analyzed extensively. Following the development of the controller, numerical simulations with generic feedback using a realistic scenario with fused vision and force feedbacks are presented to demonstrate the positioning performance capability in the nanomanipulation applications considered here.