Classification of orbital objects by spectral information

The problem of classification of orbital space objects by their spectral characteristics is considered. Analysis of the spectra and phase characteristics of materials widely used in the space industry is made. Using the principal component method, it is possible to reduce the dimension of the original feature space and separate the spectra of various materials in the new space. The same result is obtained using cluster analysis. The problem of finding the optimal parameters of the photometric system is posed and numerically solved. The result is an algorithm for determining the parameters of the photometric system and a table of the results of this algorithm for systems of 2, 3 and 5 light filters.