Research On Hysteresis Model And Anti-hysteresis Driving Algorithm Of Piezoelectric Ceramics

Piezoelectric ceramics,as the core material of piezoelectric systems,assμmed an indispensable role in high-precision and high-resolution precision manufacturing,aerospace and other fields,but their inherent hysteresis nonlinearity could bring non-negligible effects on the displacement output.Therefore,in order to improve the linear output accuracy of the system,the study of anti-hysteresis driving algorithm for piezoelectric systems was one of the important initiatives with high research value and good application prospects.This study took piezoelectric actuator as a specific research object,and built an accurate and reliable piezoelectric actuator driving system by high precision devices such as piezoelectric drive power supply,air floating vibration isolation platform and inductive micrometer,and proposed neural network model algorithm,linear equation system algorithm,binary primary linear algorithm and multi-stroke anti-hysteresis algorithm for single-stroke and multi-stroke hysteresis loops of hysteresis model respectively,and verified the feasibility of the algorithms through experiments.In addition,this study developed a multifunctional algorithmic control interface and extended the binary primary linear algorithm to the circuit module at the theoretical level.Three compensation algorithms were proposed for the hysteresis-resistant driving algorithm of the single-stroke hysteresis loop of the piezoelectric actuator.One was a dual-input neural network model algorithm based on the inverse cut rate data pair with uniqueness in the trajectory,which avoided the tedious step of repeatedly creating a network to distinguish the trajectory,and the anti-hysteresis effect was 95.78% at the highest,but the network parameter identification process was more complicated.Second,in order to reduced the complexity of the model,the linear equation system algorithm was constructed by describing the linear relationship between the cutting-rate coefficient and the sampling voltage through the partitioned segments,and the algorithm solution process was simpler,the algorithm expression was simpler,and the highest anti-lag effect was 91.16%,but due to the limitation of the fitted segments,the algorithm had a large fitting error;third,in order to reduce the impact of the fitting error,the parameters of the algorithm expression were reduced,the algorithm was constructed by describing the linear relationship between the voltage coefficient and the cutting rate coefficient,and the highest anti-hysteresis effect was 91.80%,and the algorithm expression was only a binary equation with fewer parameters and easier to obtain.In the multi-stroke hysteresis-resistant driving algorithm for piezoelectric actuators,in order to avoid repeating the tedious steps of solving the above algorithm due to hysteresis loop changes.Based on the mathematical model to describe the descent trajectory of the inverse model,the fitting relationship between the stroke voltage and the model parameters was found,and the accuracy and feasibility of the model for arbitrary strokes were checked by known and unknown strokes,with the highest anti-hysteresis effect of 93.81%The algorithm proposed in this study realized the anti-compensation control of different stroke hysteresis loops of piezoelectric actuators,and there was a more obvious progressive optimization relationship between the algorithms,and the overall anti-hysteresis effect was good,which provided a scientific reference for the study of anti-hysteresis drive algorithms.