| Summary: | 碩士 === 國立臺北科技大學 === 機電整合研究所 === 103 === The study is based on Atomic Force Microscopy (AFM) as the main instrument. Use DLC (Diamond Like Carbon) film specimen visualization technique of nano-oxidation. This research can be divided into two stages, first DLC film specimen of the oxide cytometry experiments. Including nano-oxide points, lines, and to explore the depth of the surface morphology associated with an equal width processing under different process parameters (such as voltage, speed) after the characteristic size. Thereby establishing a working database; finally, the experimental data into multiple regression analysis (MRA) with back-propagation neural network (BPN) computed. According to the results were compared with actual experimental data to verify.
During the algorithms, the first step is input data must normalize. And then the MRA with BPN were import for simulation training. The results of the anti-scale compare with experimental data for the validation error. From verify predictions show, BPN obvious to the accurate than MRA. The next step is the prediction data to compose two-dimensional graphics. And then construct the Atomic Force Microscope nano surface morphology after processing and 3D visualization of forecasting the result of the assessment.
|
|---|
