International Journal of Information Technology & Computer Science ( IJITCS )
The key condition for reliable work of electric power systems is the presence of efficient system forecasting of state variables (load flows, power flow, voltage magnitude, etc.,). Development of the state-of-the-art technique for robust forecasting of behavior of nonlinear and non-stationary power systems is one of the challenges in energetics. This paper aims to evaluate Data mining type of models for short-term forecasting of power system operating condition. Models that are examined include artificial neural networks, support vector machines; autoregressive integrated moving average and exponential smoothing. Evaluation results are presented for voltage magnitude forecasts.