Comparative analysis of electrical load forecasting models of industrial plants

The paper presents application of STATISTICA v6.0 and STATISTICA NEURAL NETWORKS software for electrical load forecasting. Relevance of forecasting is influenced by the fact that extraction of minerals in oil and gas industry is increasing. As oil extraction and transportation is very power intensive, the problem of load growth has arisen. Then, a task for forecasting of load growth occurs. The results of performed investigations show that the accuracy of short-term load forecasting using models of artificial neural networks (ANN) is better than in the case of using autoregressive integrated moving average (ARIMA) models and gives the least forecast error.