NORMALIZED ENTROPY OF HOT STARS

Abstract

In the present work the possibility of classification of the hot stars by spectra, according to their non-linear (entropy and the generalized metric) characteristics is shown. The algorithms and method of determining the information entropy from observed spectra are obtained. In this work the task was to describe quantitatively the spectra of various types on the basis of information-entropy analysis. For the research, main sequence stars and hot stars of the FSCMa type (binary systems) were chosen. A question was about the quantitative description of the difference in a spectrum of selected stars. That method of quantitative evaluation was used in various studies on normal stars. However, there are a number of algorithmic problems that we will solve related on this problem. As a result, the dependence of the normalized information entropy of the spectra on the corresponding values of the metric characteristic was constructed. The normalized entropy of FSCMa type stars belongs to a self-similarity and self-affinity regions, these objects are self-organized and have a complex, chaotic structure. In physical terms, binary systems should refer to self-organized systems, which corresponds to the proposed theoretical classification. Thus, the results of this paper show that the information-entropy analysis and calculation of the generalized-metric characteristic for spectra give us a possibility to quantitatively classify a hot stars.

Keywords: Information, entropy, hot stars, the metric – topological characteristic