Real-time identification of piezoelectric actuator nonlinearities with application to precision trajectory control

In the absence of profound understanding of the underlying physics of piezoelectric actuators, real-time identification of the system dynamics is a promising strategy for the precision trajectory control. In this paper, an on-line estimation strategy is presented for the actuator nonlinearities and ever-present unmodeled dynamics. More specifically, the influence of unknown terms is gathered into a single perturbation vector which is quantified utilizing actuator input/output real-time values. It is experimentally shown that this technique can accurately estimate actuator nonlinearities in low frequency operation, where the effect of feedback noise is not significant. Utilizing Lyapunov method and the sliding mode control strategy, the control force acting on the actuator is then designed such that the high frequency tracking control and the asymptotic stability of the system are attained. Simulation results are provided to demonstrate the effectiveness of the proposed controller for high frequency applications