Cancellation of Periodic Forces with Solid-State Actuators

Abstract

Due to their specific properties (high forces and speed), new solid-state actuators with piezoelectric or magneto-strictive transducers can be used in active damping systems to realize mechanical structures with controllable stiffnesses. One possible application is the damping of undesired periodic vibrations produced by rotating engines. The vibrations can be measured by an accelerometer, and a digital signal processing unit calculates the appropriate actuating signal to control the mechanical structure. The authors present a new procedure that uses a very simple recursive short-time autocorrelation (ACF) algorithm and a least mean squares (LMS) algorithm to calculate the control signal. The short-time ACF determines the periodicity of the vibration signal. It uses an exponentially decreasing weight function. The time constant of this function can easily be adapted to the dynamics behaviour of the mechanical process. The LMS algorithm performs the synthesis of the actuating signal. It adapts the amplitude and phase of a harmonic signal with the determined frequency. The stability of the system can be verified by a maximum adaptation rate criteria. The procedure still works with nonlinear transfer functions of the actuator, mechanics or sensor. Experiments with a magnetostrictive actuator showed a damping of the acceleration amplitude of up to -37 d B.