Using probabilistic neural network to analyze the binary stars Schulte 3, EY Cep, HD 101131, and Haro 1-14c

Abstract

The use of artificial neural networks (ANNs) in physical sciences has increased recently. Determining the orbital elements of binary systems helps us to obtain fundamental information. In this paper, ANNs were used to find the corresponding orbital and spectroscopic elements of four double-lined spectroscopic binary stars: Schulte 3, EY Cep, HD 101131, and Haro 1- 14c. The orbital parameters of the radial velocity curve obtained from ANNs were compared with other traditional methods and we show that the proposed method is of high accuracy. Our numerical results are in good agreement with those obtained by others using nonlinear regression methods. We show the validity of our new method in a wide range of different types of binary. In this method, the time consumed is considerably less than in the other traditional methods. The present method is applicable to orbits of all eccentricities and inclination angles and enables one to vary all of the unknown parameters simultaneously.

Keywords

• probabilistic neural network
• binary systems
• eclipsing
• velocity curve