Control of MEMS on the edge of instability

This paper investigates a new method for tracking parameters at which dynamic bifurcations occur in MEMS based on state statistics of a resonator. Experimental results show that observed changes in phase and amplitude precede a sub-critical pitchfork bifurcation. Feedback control, based on the statistics of the phase, is then employed to stabilize a device on the edge of instability. Sensitivity to changes in device parameters is explored and initial results show high sensitivity is obtained for low vibration amplitudes. Furthermore, experimental results on the same device using an older method of bifurcation sensing show acquisition rate improvement by over three orders of magnitude with the new method.