Applicability of models based on deterministic chaos theory for forecasting of electricity sales in rural areas

The applicability of models based on deterministic chaos theory, in particular the self-similar fractal models, logistic models according to Prigogine, logistic models according to Schuster, and heuristic crossing models for one-year-ahead forecasting of the sales of electric energy in the rural areas, is verified in the paper. The model quality assessment was carried out based on the errors analysis of the 10-year ex post forecasts of the electric energy sale to rural consumers by three selected distribution companies. As the electric energy consumption were disturbed by economic crises, the forecasts were developed using not only annual sales value but also using their 2-, 3-, 4- and 5-year totals. The best forecast were obtained when the electric energy sales were predicted using the Schuster and heuristic crossing models, made on the basis of four- and five-year totals of annual sales values.