Assessment of technical losses of electrical energy in distribution networks based on artificial neural networks

Abstract

Calculation of electrical energy losses is one of the difficult problems in the process of transmission and distribution of electricity. At present, for calculating electrical energy losses, some methods based on deterministic information are used, which involves the adoption of a number of assumptions and using the results of iterative steady-state calculations. For a quick and correct calculation of losses, it is necessary to use methods that allow accounting for the fuzziness of circuit-mode information, as well as faster approaches required for the operational management of electrical networks. Considering this, the article presents models for assessing electrical energy losses in distribution networks based on various types of artificial neural networks. The use of these models makes it possible to increase the speed of calculations of electrical energy losses. Approbation of theoretical statements is carried out on the example of real circuits of 6-10 kV distribution networks.