Comparison of various methods for predicting electricity losses in networks

Abstract

Relevance: electricity losses in 0.4 kV distribution networks are an urgent problem that requires a solution.  This problem is especially acute in conditions of increasing load on networks associated with increased electricity consumption and the introduction of new technologies. Traditional loss forecasting methods, based on statistical models and empirical data, are often not accurate enough and do not take into account the specifics of modern networks. In this regard, there is a need to use more advanced methods that would provide an accurate and reliable forecast of losses. This article discusses the use of the method of group accounting of (GMDH) arguments to solve the problem of predicting electricity losses in 0.4 kV networks.

Aim: application of the GMDH method in the development of an algorithm and model for the analysis, assessment and forecasting of electricity losses in 0.4 kV electrical distribution networks, assessment of the accuracy and efficiency of the proposed model.

Methods: There is a whole arsenal of forecasting methods, each of which has its own strengths and weaknesses. The most frequently used methods are regression analysis, ARIMA, GMDH, ANN and others. Selecting the optimal method is a key factor for successfully predicting power losses.  This will allow us to develop effective strategies to reduce losses and achieve economic and environmental feasibility of the energy system.

Results: a model for analysis, assessment and forecasting of electricity losses in 0.4 kV electrical distribution networks is substantiated and proposed.