Hysteresis Modeling And Control Simulation Based On Gaint Magnetostrictive Materials

Giant Magnetostrictive Material(GMM),as a helpful kind of intelligent material,has the advantages of short response time,large stroke and large loading force,compared with common piezoelectric ceramics.In modern industrial field and scientific research,nanoscale precision positioning technology has better application prospect in microelectromechanical field and optical field.With a mount of rare earth,China has a high level in the design of giant magnetostrictive actuator,but the development of control is slow.In this paper,we design a feedforward inverse compensation and adaptive fuzzy PID control for giant magnetostrictive actuator to improve the control accuracy of the actuator.This thesis firstly introduces the principles and properties of super magnetostrictive materials,then builds a data acquisition platform to capture the corresponding displacement changes by controlling the voltage intervals and produces a table of F-functions,which is used as the training data set for the neural network model in Chapter 3.The hysteresis nonlinearity is investigated by building a hysteresis model using the Preisach operator and obtaining an extended Preisach model by improving the model.A neural network is built using Tensor Flow to identify the input and output data of the extended Preisach model and a positive model of the super magnetostrictive actuator is obtained.In order to achieve the erasure properties of the Preisach operator,further algorithms are designed to improve the neural network model.The model is proved to be more accurate at displacements greater than 3?m by experimental simulations.By analysing the positive model,the Preisach inverse model is also constructed using a neural network,and the inverse model is connected in series with the positive model to form a feedforward control to compensate for the hysteresis non-linearity.Through experimental simulation,the feedforward inverse compensation can eliminate most of the non-linear errors.In order to further eliminate the error of feedforward control,this paper proposes a composite control method combining feedforward control and fuzzy adaptation PID control.After analysis of the simulation results,the proposed composite control scheme further improves the control accuracy of the control system on the basis of the feedforward control scheme,with an average error of 0.2268?m and a percentage error of 0.22%.The work done above provides a theoretical basis for the design of high precision displacement tracking controlled giant magnetostrictive actuators in precision displacement technology.