Research On Displacement Tracking Control Strategy Of Piezoelectric Driven AFM

With the continuous development of nanotechnology,higher requirements have been put forward for the detection and processing of nanoscale.Atomic force microscope(AFM),as an experimental instrument with nanoscale detection and nanoscale operation function,has a complex system,whose performance is affected by hardware equipment,control algorithm,imaging strategy and other factors.Piezoelectric ceramic actuator is the key component of atomic force microscope,which directly determines the resolution performance of atomic force microscope.In terms of this issue,thesis will carry out the following research:Firstly,consult the information about the unique hysteresis and nonlinearity of the piezoelectric ceramic driver,and construct the mathematical model of the piezoelectric ceramic driver according to the unique hysteresis characteristics of the piezoelectric ceramic driver.In the mathematical model,a simplified PI model,As the main model of the feedforward controller,the PI inverse model is as close as possible to the real hysteresis inverse model will be more helpful to the final closed-loop control effect,but because the asymmetric model is difficult to invert the model and the inverse of the asymmetric model The existence of the model is in doubt,and a symmetrical PI inverse model close to the reality will be established to achieve the purpose of feedforward compensation.Secondly,in order to ensure the accuracy and performance of AFM,the controller is designed and verified.Based on the analysis and research of the current AFM control system,the PID pure feedback control and separate feedforward compensation control,separate ADRC control and ADRC feedforward compensation control are designed respectively,and the above control methods are compared and verified.Because the above control methods have shortcomings in control accuracy and real-time performance,BP neural network is introduced to optimize on this basis,and nonlinear BPNN-ADRC and feedforward compensation type BPNN-ADRC are designed,and they are applied to atomic force microscopes.In the control system,AFM can obtain the ability of self-learning,thereby improving the accuracy and real-time performance of the system.Finally,through Matlab/Simulink simulation verification,compared with the above control methods,the simulation results show that the feedforward compensation BPNN-ADRC has certain advantages in displacement tracking control.