关键词: 电力需求预测, 非线性回归组合神经网络, 二重趋势性

Abstract: Power demand possesses dual property of increasement and seasonal fluctuation simultaneously，so it makes power demand variation possess complex non-linear combined character.To improve the forecasting accuracy of power demand，a new forecasting model which named non-linear combined neural network is put forward.The model can effectively use the advantages of non-linear regression analysis and artifical neural network and improve the forecasting accuracy of power demand obviously.The simulation result indicates that the model is effective and feasible for forecasting power demand.At the same time，the model is an effective tool to solve the problems of forecasting for other similar seasonal time series.

Key words: power demand forecasting, non-linear regression combined neural network, double trend