Mechanical Model Analysis Of Nanomanipulation And Predictive Control In SEM

Nanomanipulation technology is a cutting-edge technology in the research field of micro and nano technology. It is related to the development and research of many fields. The research of adhesion is the key issue during nanomanipulation. Micro-environment is different from the macro-environment.Due to the influence of size effect, adhesion as the main influence force. In order to avoid bending and fracture phenomenon of nanowire in nanomanipulation, this paper established and analyzed the nanowires force model in SEM nanomanipulation; The predictive control system was introduced to ensure the operation of good accuracy and real-time during nanomanipulation, and the BP neural network prediction model was used to control the nano positioner Attocube.In order to study the adhesion models between the nano components in nanomanipulation, on the basis of mechanical model analysis in Nano environment, the adhesion models between particles had been established, and on this basis, the adhesion model of nanowire-nanowire, nanowires-substrate and probe-substrate had been established, then the relationship between adhesion and the distance of components has been analyzed.For the study of nanowires bend or fracture phenomenon in the process of nanomanipulation, first of all, the mechanical properties indicators of nanowires were analyzed, such as bending stress and fracture strength, then combined with the adhesion phenomenon theory, the relationship between the length and radius of the bending nanowire has been analyzed. At last, the Matlab simulation software and Lammps molecular dynamics simulation analysis software were used to simulate analysis.To achieve an automation nanomanipulation control system, the predictive control system based on nano positioner Attocube has been designed. The controlsystem adopts BP neural network predictive control algorithm, and the predictive model and the optimization model was both used two rolling three-layer neural network. Firstly, the Approximation effects of BP neural network algorithm and the prediction effect of predictive control algorithm was both verified. Secondly,the math model of Attocube has been built, then, the model of the whole control system has been established. Thirdly, the step signal and sine signal respectively as an input signal to simulate the two working models of Attocube. The predictive control of the whole process of nanomanipulation was completed.Predictive control theory has been introduced to the nanomanipulation control system in SEM, and the simulation results show that predictive control can be applied to micro and nano operation, and better control can be carried out more in-depth study.