Research On Motion Control Of Atomic Force Microscope Based On Self-Tuning Fractional Order PI~?D~?

With the continuous development of nanotechnology,the nanoscale detection and processing put forward higher request,atomic force microscope(AFM),as an experimental apparatus with detection and nano operating functions,because of its small size,high precision,low manufacture cost and easy promotion,etc are widely used in physics,chemistry,biology,materials,electronics and other fields.As a complex system,the atomic force microscope's performance is affected by factors such as hardware equipment,control algorithms,and imaging strategies.At present,the performance of the AFM system cannot meet the performance requirements in highprecision motion scenarios.The traditional PID control has limitations due to the algorithm itself.Sex has become the main factor limiting the performance of AFM systems.In order to improve the control accuracy,a two-degree-of-freedom fractional order controller is designed to solve the problems of insufficient control performance and complicated parameter setting in traditional PID control.On this basis,the BP neural network is introduced to realize the controller parameter self-tuning.The main research contents of this article are as follows:1.Through the design of the semi-physical simulation platform based on xPC target,the system characteristics and existing control problems of the atomic force microscope are studied,and the hysteretic nonlinear characteristics of the piezoelectric ceramic actuator are modeled.2.Aiming at the problem of insufficient control accuracy of AFM system under traditional integer order PID control,the AFM motion control method based on twodegree-of-freedom fractional order is studied.For the problem that the fractional differential operator in the controller cannot be directly digitally implemented,the Oustaloup approximation method is used to approximate the discretization process,and the influence of the controller parameters on the performance of the AFM system is analyzed according to the established frequency domain transfer function model of the AFM system,And gives the controller parameter tuning method of the system phase margin as the main index.3.Aiming at the complicated parameter setting of two-degree-of-freedom fractional order controllers,the AFM motion control method based on BP neural network and two-degree-of-freedom fractional order is studied,which realizes the self-tuning of controller parameters and improves the control performance.The two-degree-of-freedom fractional order part of the controller is designed to reduce the fractional order calculus operation of the BP neural network of the controller.Design the BP neural network part of the controller to make the structure of the BP neural network as simple as possible under the premise of ensuring the real-time performance and error accuracy of the system control,and optimize the parameters of the remaining parameters of the controller.4.The designed controller is applied to the atomic force microscope system,and the trajectory tracking,sample imaging and sample characterization experiments of the AFM system are realized on the semi-physical simulation platform based on xPC target,which proves the effectiveness of the controller designed in this paper.The results show that the controllers designed in this paper are superior to the traditional PID controllers in terms of performance,which improves the system performance of the atomic force microscope.