| Infrared optical materials are widely applied to the photoelectron precisionequipment, infrared cameras, large-scale laser windows, semiconductor wafers, but atpresent most of light source of laser heterodyne interferometer in extensive use is thevisible light, and their work patterns are basically reflective. Therefore they can notmeasure precisely the internal microstructure of infrared optical materials such asoptical thickness, internal stress and so on. In this article, MEMS scanning infrared laserheterodyne interferometer with the wavelength of1550nm in use is able to have aprecise measurement of the internal microstructure of infrared optical materials becauseits light source is in near-infrared region, and the operating mode is transmissive.MEMS scanning infrared laser heterodyne interferometer is a latest achievementthat combines the micro-electromechanical system with traditional laser heterodyneinterferometer. It not only possesses the advantages of high measurement speed, highaccuracy, strong anti-interference ability, good reproducibility and traceability as theordinary heterodyne interferometer, but also holds very good dynamic measurementcharacteristics perfectly that suit for online measurement. This article mainly aims toprovide a set of high-precision data processing algorithms in order to meet therequirements of the data processing in the heterodyne interferometer. The algorithm canbe divided into two parts: firstly, to extract the information of phase difference to get thediscrete thickness data of the object under test; secondly, to gain the thickness undulatedata of the real object under test according to the data of the first part, thus referring tothe problem of data fitting.In the first place, it is introduced of the working principle of MEMS scanninginfrared laser heterodyne interferometer with the wavelength of1550nm on accuratelymeasuring the optical homogeneity and the relative optical thickness and put forward ahigh precision processing algorithm for the data of the heterodyne interferometer, i.e.the phase extraction and surface shape fitting algorithm for MEMS laser heterodyneinterferometer in research.Then, three kinds of algorithms for extracting the decimal part phase difference areput forward: FFT method, IQ orthogonal demodulation method and least square method.And it is analyzed and discussed in detail respectively on the three algorithms throughthe signal-to-noise ratio of intermediate frequency, the sampling frequency andsampling depth of the A/D data acquisition card, and other factors that affect theprecision of phase extraction.Furthermore, concerning the data fitting part it is proposed to use the method ofZernike polynomial fitting wavefront interferometer. In consideration of the high-volume of data in this article, it is proposed to apply the Gram-Schmidt orthogonalmethod and the Householder transformation method to solving the Zernike polynomialfitting wavefront interferometer. Thus the simulations are presented on thecomputational accuracy of the two methods by changing the order of the Zernikepolynomials.Finally, least squares phase extraction algorithm and Zernike polynomial fittingalgorithm are used together to form the phase extraction and surface shape fittingalgorithm for MEMS laser heterodyne interferometer. Then the stimulation experimentis taken based on the system parameters and the phase extraction accuracy of thealgorithm in stimulation is9.02105radians and data fitting computation accuracy0.0046λ. Lastly the algorithm is applied to measure the decimal part relative thicknessof the heterodyne interferometer system in no-load. |
PDF Full Text Request