Modeling the dead-band in magnetic actuation

We present a simple model and a controller based on this model for non-contact actuation. A magnetically driven micro tool (MMT), often embedded inside a microfluidic chip, can be driven by a magnet which can be positioned in micro order accuracy using linear motors. Position errors caused by friction between the micro tool and surface can be modeled. After calibration of the model parameters using encoder data from linear motors, and vision feedback, this model can be used as a controller. Proposed controller is applied to a microfluidic chip and an accuracy of 12μm is achieved which is 3 times better than previous controller.