Adaptive levitation control using single neuron for micromachined electrostatically suspended gyroscope

摘要

An efficiently adaptive single neuron proportional integral (SNPI) controller for a non-silicon micromachined electrostatically suspended gyroscope is proposed and experimentally validated. The microgyro-scope utilises a spinning and five degree-of-freedom suspended rotor as the proof mass, and thus is characterised by potential navigation-grade precision. The SNPI controller, which features simplicity, efficiency and robustness, is explored to realise reliable initial levitation and stable suspension of the rotor in the presence of dynamic nonlinea-rities, microfabrication imperfections and disturbances. Experimental results demonstrate that a desirable initial levitation process can be achieved with 10% overshoot at most and 15 ms setting time using the SNPI controller, which also works well in response to a 10 Hz, 3 Vp-p mimic square axial acceleration input.