Autonomous vibration suppression using on-line pole-zero identification

An on-line identification and control algorithm is developed based on the p roperties of collocated sensing and actuation. The feedback control law con sists of second-order compensators that achieve equivalent damping in both the filter dynamics and resonant structural dynamics, thus maximizing the d amping in the structure and controller Optimal design of the feedback compe nsator is obtained using a pole placement algorithm applied to a single, un damped resonant mode. Numerical analysis indicates that multiple modes and structural damping do not appreciably change the damping achieved using the optimal parameters. The pole placement analysis demonstrates that only the pole-zero spacing and DC gain of the collocated transfer function are requ ired to choose the optimal parameters. An on-line identification procedure is developed that sequentially determines the DC gain and pole-zero spacing and automatically designs the feedback compensator. This forms the basis f or the autonomous control algorithm. Experimental results on a flexible bea m demonstrate that the procedure can accurately identify the pole-zero spac ing and automatically design the feedback compensator. A fivefold increase in damping is achieved in the first mode and a twofold increase in damping is achieved in the second mode. Discrepancies between predicted and measure d damping are attributed to phase lags due to signal conditioning and low-p ass filtering of the sensor signal.