Empirical verification of certain method for forecasting non-stationary processes.

The paper presents an empirical verification of the method for forecasting non-stationary processes. The forecast and the estimation of its error are obtained as a result of solving two forecast parameter estimation problems. The verification is presented by means of such a real process as the power system load. The analysis of predicting the quality changes, depending on the changes in values of its determinants, is carried out. The analysis proves that the prediction method being verified yields good results - the forecasts computed based on it are reliable, because their errors do not exceed 3%.