关键词: 灰色-小波网络, 短期电力负荷预测, 组合模型, GM（1, 1）预测

Abstract: Short-term electric power load data have the traits of fluctation and discretization.The paper uses grey forecasting to reduce fluctuation of original load data and combines it with the data of grey forecasting to get new series，and thus takes the new series as training samples to put forward a new method for forecasting short-term electric power load based on grey-wavelet network combinational model.The new method combines the advantages of grey theory，wavelet analysis and artificial neural network effectively.Compared with BP network，its speed of convergence and precision are much higher than BP network.The results shown by the simulation indicate that the method is effective and feasible for forecasting short-term electric power load.

Key words: grey-wavelet network, short-term electric power load forecasting, combinational model, GM（1, 1） forecasting