Self-optimizing and adaptive digital signal processor based algorithms in scanning tunneling microscopy

A self-optimizing proportional integral differential feedback loop and an a daptive scan generator have been implemented in a digital signal processor board EVM56k from Motorola for the control of a scanning tunneling microsco pe. The self-optimizing feedback loop uses an evolutionary algorithm to opt imize the step response of the system. The variation of the target step res ponse strongly determines the results of the optimizing process and allows to optimize either for highest scan speed or best signal-to-noise ratio. In combination with the improved adaptive scan generator, the adjustment of t he scan velocity to the local topography of the sample increases image cont rast and quality considerably. (C) 1998 American Institute of Physics. [S00 34-6748(98)00312-8].