Iterative Control Method Of Piezoelectric Ceramic-Driven Micro-Positioning Platform

As a new type of functional material,piezoelectric ceramic actuators have the ad-vantages of low energy consumption,accurate positioning,and high reliability.They are widely used in mechanical micro-operation and vibration control.However,there are characteristics such as multi-value mapping and frequency dependence at the same time,which will cause the decrease of system control accuracy.Therefore,it is of great practical significance to study how to eliminate the influence of the frequency-dependent nonlinear characteristics of piezoelectric ceramics and improve its tracking accuracy.Based on the constructed piezoelectric ceramic drive experiment platform,a Ham-merstein model that can describe piezoelectric ceramic actuators is established.The static nonlinear part of the Hammerstein model adopts the asymmetric Bouc-Wen model,and its parameters are identified by the differential evolution algorithm;the linear dynamic part of the Hammerstein model is equivalent to a linear system,and the parameters are identified by the least square method in the matlab identification toolbox.The built Hammerstein model verifies its accuracy through voltage drive experiments.In order to eliminate the hysteresis characteristics of piezoelectric ceramics,a strategy of inverse compensation and linearization of the controlled object is adopted.Therefore,based on the Hammerstein model,a strategy of hysteresis compensation combined with feedback-iterative learning control is proposed to deal with the nonlinearity and uncer-tainty of the system,in which hysteresis compensation offsets the hysteresis nonlinearity of piezoelectric ceramics;iterative learning control is essentially It is a feed-forward control that can deal with repeated interference and hysteresis compensation errors after a limited number of iterations to improve control accuracy;feedback control handles external inter,ference and improves system stability.The experimental results show the effectiveness of the designed compound control under the expected displacements of different frequencies.In the actual control process,in order to avoid the existence of various interference sources in the environment and the noise generated inside the experimental platform that will affect the estimation of the system state,the linear Kalman filter method is introduced on the basis of hysteresis compensation;at the same time,taking into account Model predictive control with a "feedforward-feedback" structure can handle the model uncertainty caused by compensation errors and the advantages of being easy to combine with Kalman filter.A model predictive control based on Kalman filter is designed.The experimental results show the effectiveness of the designed control strategy under the expected displacements of different frequencies.