Stochastic Dynamic Trapping in Robotic Manipulation of Micro-Objects Using Optical Tweezers

Various automatic manipulation techniques have been developed for manipulating micro-objects using optical tweezers. Because of the small trapping force of optical traps and increase in kinetic energy during manipulation, a trapped object may not remain trappable, especially in the presence of random Brownian perturbation. However, there is no theoretical analysis so far to help understand the effects of dynamic motion and Brownian forces on the trappability problem of optical tweezers. This paper investigates the optical manipulation of micro-objects under random perturbations. Here, we provide for the first time a theoretical and experimental analysis of the dynamic trapping problem from stochastic perspectives. We derive the relationship between trapping probability and maximum manipulation velocity. A controller with appropriate velocity bound is then proposed to ensure that the system is bound and stable. The experimental results confirm the accuracy of our theoretical analysis and illustrate the necessity and usefulness of the proposed controller.