Multivariable control of a metrological Atomic Force Microscope

Atomic Fore Microscopes (AFMs) are widely used for nanoscale applications. Until recent developments these instruments were controlled in open loop or by a PID type controller. In the last decade model-based control techniques emerged for controlling AFMs, but in almost all cases the multivariable behavior was ignored. In this paper, we present a multivariable control strategy using the standard plant framework, applied to a metrological AFM, used for the calibration of transfer standards. Although this device has a low amount of coupling, we will show that multivariable control has advantages. By using multivariable control, the amount of coupling is reduced compared to decentralized control. Simulations show a significant reduction of the errors in x, y and z directions by applying multivariable control.