Random vibration control based on adaptive filter

Abstract

The random vibration adaptive control method is presented based on the recursive least-squares (RLS) adaptive filter. The power spectral density (PSD) of drive signal is updated based on the off-line identification of frequency response function (FRF) in conventional random vibration control system. And the pseudo-random drive signal generated by the conventional time-frequency transformation has leakage between frequency resolutions. The control precision is difficult to meet with conventional random vibration control method. The finite impulse response (FIR) filter is designed with the reference PSD in random vibration adaptive control system. And the time domain drive signal is generated by filtering a series of independent white noise. The RLS adaptive filter is designed to obtain the inverse model of the device under test (DUT) based on adaptive inverse control algorithm. The time domain drive signal is filtered by the inverse model so the PSD of the DUT output signal can replicate the reference PSD in high precision. Random vibration test results show the effectiveness of the proposed algorithm.