EMD-Based Short Term Electric Power Load Forecasting

The forecasting for the load of electric power system is fundamental to ensure safe and economic operation of power system. For a power system, safe and economic operation of power grid as well as good electric quality depends on correct estimation of electric load.How to select one of the forecasting techniques is the primary problem of the load forecasting, namely,how to build a forecasting model which is applicable to the studied district. Along with the development of the technology, the study of STLF has been deepened and various methods have been put forward. Because of the advanced method from the classical Statistic analysis methods to the modern intelligent methods, the accuracy has been improved continuously. However, they have some shortages inevitably, due to the complexity of the forecasting problem.Time-frequency analysis method based on Empirical Mode Decomposition (EMD) is a new analysis method of signals. After more than ten years of development, The method of theoretical research value and broad application prospect has gradually been recognised, it handles the non-stationary signal ability is taken seriously by people place. The definitions of instantaneous frequency, characteristic time scale and intrinsic mode function are illustrated; and mainly studied the EMD decomposition principle and algorithm.A new method for short-term load forecasting is presented in this paper. Based on the characters of power load series such as periodic randomness and not stationary, firstly, the power load series is decomposed into intrinsic mode functions (IMF), then, According to the characters of each IMF, using System Identification Toolbox of Mat lab 7.0 to select models (AR,MA,ARMA),set orders,realize forecasting. At last reconstruct the predictive value. The simulation is carried out with the data of EUNITE competition and the results show that this method has rather high precision in forecasting.