Hysteresis Model And Control Of Piezoelectric Ceramics Based On Improved Neural Network

Precision positioning technology is widely used in microscope technology,bioengineering technology,semiconductor technology and other fields,and its rapid development can improve the level of national advanced manufacturing technology to a certain extent.At present,there is still a big gap between the research level of precision positioning technology in China and the top level in the world,so it is particularly necessary to study precision positioning technology.As one of the executive carriers of precision positioning technology,piezoelectric ceramic actuator has been widely used in precision positioning system because of its fast response speed,large output force,high resolution,no noise and so on.Due to some inherent properties of piezoelectric materials,piezoelectric ceramics show some complex characteristics in displacement output,such as hysteresis non-linearity,creep characteristics,dynamic characteristics and etc.,which seriously affect the accuracy of positioning.Therefore,the micro-motion platform driven by a piezoelectric ceramic actuator is the research object in this paper.Based on the analysis of the nonlinear characteristics of piezoelectric ceramics,a rate-independent hysteresis model of BP neural network based on adaptive particle swarm optimization is established.Based on the analysis of the dynamic characteristics of piezoelectric,a rate-dependent hysteresis model of piezoelectric ceramics based on analog filter is established.In order to improve the control accuracy,the feed-forward open loop control based on inverse model and the composite controller based on the rate-dependent hysteresis model are designed.The effectiveness of the controller is verified by experimental research.The detailed contents of the research are as follows:(1)The experimental platform for measuring the input and output characteristics of piezoelectric ceramics is set up,and the complex hysteretic characteristics of piezoelectric,including creep characteristics,dynamic characteristics and asymmetries in hysteretic characteristics,the congruent minor loop property and the wiping out property,are studied experimentally.(2)a rate-independent hysteresis model of neural network based on adaptive particle swarm optimization(APSO)is established.Based on the basis of BP neural network,adaptive particle swarm optimization algorithm is introduced to solve the problem that neural network is easy to fall into the local optimal solution.The simulation results show that the model can fully describe the hysteresis nonlinear,and compared with the other two kinds of lag models,its superiority is proved.which aims at solving the problem that BP neural network easily fail into local optimal solution,introduce an adaptive particle swarm optimization algorithm based on BP neural network.The results of simulation experiments show that the model can fully describe the delay nonlinear,and compared with the other two hysteretic models,its superiority is proved.(3)According to the dynamic variation law of piezoelectric micro-motion platform,a second-order low-pass analog filter is introduced on the basis of the rate-independent model to describe the dynamic characteristics of piezoelectric micro-table.The simulation results show that the accuracy of this method is obviously higher than that of static model under different frequency inputs.(4)The inverse compensation control method of piezoelectric micro-platform is studied.Using the inverse model,the feed-forward controller based on the rate-independent inverse model and the rate-dependent inverse model and the composite controller combining the rate-dependent inverse model with the neural network PID are designed respectively.Finally,the effect of the designed model is verified by tracking experiments.