Modeling And Inverse Compensation Control Of Actuators With Hysteresis

Piezoelectric actuator(PZT)and giant magnetostrictive actuator(GMA)are widely used.However,their inherent hysteresis effect and complex nonlinear characteristics usually bring instability to the system.The modeling and control of smart structures has important theoretical significance and engineering application value.To eliminate the influence of the hysteresis nonlinearity on the control precision,this dissertation dose a deep research on the theory and method of the rate dependent hysteresis nonlinear system,and also achieves the real-time tracking control of PZT and GMA.The main research contents and achievements are listed as follows.Hammerstein model of PZT and GMA are established.BP neural network and ARX model are used to characterize the hysteresis nonlinearity and rate dependence of the actuators.The extended space input method is used to overcome the multi-value mapping of actuators.The modeling results show that the designed model can describe their hysteresis nonlinearity whether it is PZT or GMA.The model is easy to identify and have strong ability of frequency generalization.The feedforward-feedback composite control strategy is designed to track the PZT and GMA.The feedforward controllers are designed based on the Hammerstein inverse model.Feedback controllers are designed based on the PID control,single neuron PID Control and fuzzy PD control respectively.The experiment system of actuators real-time tracking control based on dSPACE platform is built.The systems are designed for PZT and GMA respectively.The experimental results show that the composite control strategy can track the PZT and GMA,which can meet the needs of engineering and research.