Assessment method of numerical integration used in measuring profile based on ultra-precise thin light beam scanning

Abstract

In order to improve the precision of profile measurement based on ultra-precise thin light beam scanning, an assessment method that compares different numerical integration algorithms in frequency-domain is put forward. The compared numerical integration methods are regarded as recursive digital filters. Through comparing their functions of frequency response in frequency-domain, the delivering role of noise with different frequencies can be analyzed directly and clearly in the process of integrating measured slope data. Analyzing results show that the method of cubic spline is better than trapezoidal, Simpson and 3/8 Simpson rules.

Keywords

Integration Method / Reconstruction Error / Spline Interpolation / Numerical Integration Method / Target Profile

References

[1] ZhaoY., LiZ., LiD., XiaoT., XiaS.. Optics Communication, 2001, 200: 23

[2] ElsterC., WeingartnerI.. Proceedings of SPIE, 2002, 4782: 12

[3] RakelsJ. H.. Nanotechnology, 1998, 9: 49

[4] PressW. H., FlanneryB. P., TeukolskyS. A., VetterlingW. T.. Numerical Recipes: The Art of Scientific Computing, 1987, Cambridge, Cambridge U. Press

[5] MathewsJ. H., FinkK. D.. Numerical Methods Using Matlab, 1999, Englewood Cliffs, N. J, Pretice-Hall