Yüksek başarımlı AKM taraması için tümleşik bir LQG ve tekrarlı kontrolcü

In order to improve the image quality at high scan speeds in AFM imaging, a combined controller architecture employing a linear quadratic Gaussian (LQG) controller and a repetitive controller (RC) is proposed. To calculate the optimal values of the parameters used in the proposed LQG+RC design minimizing the scan error, a Matlab/SIMULINK model of the whole scan process is developed first and then genetic algorithms are employed on the simulation model. In our design, the LQG controller provides more robustness and better tracking performance than a traditional PI controller used for the same purpose. Moreover, since the profile of a scan line is treated as a disturbance in our approach, the proposed LQG controller handles the large variations in scan profile better than the PI controller. The RC on the other hand extends the controller bandwidth and further improves the tracking performance. Our simulations and the experimental results show that the scan performance of the proposed LQG+RC design is better than that of the PI+RC design implemented in our earlier studies.