A parametric approach to the prediction of the time-behavior of harmonic-quantities in electrical networks

In this paper, the problem of harmonic modeling and prediction is dealt with the time-series approach. Techniques for order and parameter identification of seasonal auto-regressive integrated moving-average, SARIMA, models, are discussed. Forecasting tools, valid for power, harmonic distortion and harmonic-current amplitudes, are obtained, which provide accurate daily predictions. Moreover, the stochastic dependence between the investigated quantities are outlined.