An atomic force microscope (AFM) has been utilized to implement various manipulations with nanometer precision. Unfortunately, an AFM-based nanomanipulation system often meets the problem of low reliability and low efficiency,
mainly due to the difficulty of positioning the AFM tip near the sample. In this study, a positioning control strategy is designed to accurately drive the probe to implement nanomanipulation tasks. Specifically, for an AFM piezo-scanner, a novel control strategy consisting of the following three algorithms are proposed to alleviate positioning error caused by such factors as piezo-hysteresis, cross-coupling and other uncertainties: (a) an image-based hysteresis compensation algorithm, which first obtains the voltage-displacement relationship of the hysteresis for the AFM piezo-scanner by analyzing some collected images for a calibration grating, and then utilizes this relationship to compute suitable control inputs to compensate for the positioning error caused by hysteresis nonlinearity; (b) a landmarkbased positioning algorithm addressing cross-coupling effect,which first indents the sample to make regular landmarks by a series of control voltages, based on which a polynomial curve fitting method is then utilized to calculate proper inputs so that cross-coupling effect can be compensated efficiently; (c) a local scannning-based compensator, which addresses the positioning
error caused by thermal drift or other uncertainties within the nanomanipulation system successfully. Some experiment results are included to show that precise nanopositioning performance can be achieved by using the presented approach.