Error synthesis and statistical analysis on stepped mirror array by Monte Carlo method

Abstract

To achieve the miniaturization and the static state of the Fourier transform spectrometer, two stepped mirror arrays are introduced into the time-modulation Fourier transform spectrometer to replace of the plane mirrors. The two stepped mirrors can sample the interferogram data in two-dimensional space, which can reduce the size of the instrument and increase the stability of the system. Due to the precision restriction on the stepped mirrors in the fabrication process, the various sub-mirrors of the stepped mirrors may contain various thickness errors and angle errors, which can affect the distribution of the interferogram and the quality of the spectrum. We regard the thickness error and the angle error of all the sub-mirrors as random variables, and synthesize all the error terms into a Fourier transform integration function using Monte Carlo method. By means of statistic analysis on the spectrum error factor, we can appraise the recovered spectrum affected by the thickness error and the angle error of the sub-mirror. The statistical result indicates that the statistical mean of the spectrum error factor increases with thickness standard deviation and angle standard deviation increasing. According to the statistical analysis on spectrum error factor, the tolerances of the thickness standard deviation and the angle standard deviation of the sub-mirror can be determined in the fabrication process of the stepped mirrors.