PHYSICS AND MATHEMATICS

On the issue of neural network modeling to assess the energy consumption of large arrays of consumers (buildings)

L. P. Haritonova

Volgograd State University of Architecture and Civil Engineering

Abstract: Research is devoted to the mathematical methods of modern applied mathematics, namely neural network modeling in the study of the processes of different physical nature, the complex analysis of the use of artificial neural networks and neural network modeling to assess the energy consumption of large arrays of consumers. Regression analysis predicted results and the target output are made. The equations of the regression lines and correlation coefficients are obtained in different geometries neural network and calculation algorithms. Subject promising for the prediction and management of the city (district, etc.) economy as a whole to avoid outages, emergencies, providing residents' needs and their comfortable living.

Keywords: neural network modeling, large arrays of consumers.

DOI: 10.18454/IRJ.2016.50.176