Forecasting of electric power losses in low voltage network based on regression analysis

Abstract

Regression analysis is one of the most widespread methods of electric losses evaluation, which allows getting a regression equation relating electric losses to generalized factors affecting them. In this scientific article, regression models have been examined to definition of the equivalent resistance of lines and electric losses in low voltage electric networks. To solve this problem factors have been selected that affect the value of the output quantity and limits of their variation have also been identified. These factors include: line length, the network load and the cross-section of lines, the number of lines and the ambient temperature. Regression models were derived by virtue of the generalized parameters according to expressions given in the article to determine the equivalent resistance of the power supply circuits. Such models allow for the dynamics of change of the guild electric network. Also nomograms were constructed for lines with different cross-section, the average length and loading, that allow to determine the equivalent resistance of any configuration guild network without any time-consuming calculations.

Keywords: electric power losses, regression analysis, the equivalent resistance of the lines.