Online controller optimization for piezoelectric hybrid micropositioning systems

The design of an efficient controller is from essential importance in many industrial applications. This is particularly true for the wide spectrum of nanopositioning tasks, where in many cases centimetre-range trajectories with the maximal positioning error held in the nanometre range has to be followed. In this paper a model-based control design approach for hybrid micropositioning systems, capable to fulfil both imposed requirements, is presented as first. Thereafter, online adaptation methods for the control parameters by changed conditions - for example for positioned objects with different mass - are given. The described control concepts are implemented on an FPGA and a CPU-based real-time system and experimentally validated using a commercial available hybrid micropositioning stage. Based on experimental results, the appropriability of various optimization methods is rated.