Time series prediction for electric power industry with the help of syntactic pattern recognition

Load prediction is one of the most important problems in electric power industry. The prediction is usually made with the help of standard time series analysis models. The novel syntactic pattern recognition-based model for the load prediction is defined in the paper. The Syntactic Pattern Recognition-based Electrical Load Prediction (SPRELP) System is described and the results concerning the reduction of the forecasting error with the comparison with other methods are presented.