基于神经网络的短期电力负荷预测及其MATLAB实现

Short-term Power Load Forecasting Based on Neural Network and MATLAB

针对短期电力负荷预测问题,利用MATLAB软件建立日最高温度、日最低温度、日平均温度和平均湿度等气象因素对电力负荷的回归预测模型,具体对比给出BP神经网络与NARX神经网络两种回归预测结果,并通过对隐含层网络参数的调试对BP神经网络进行了适当的改进.

For short-term power load forecasting problem, using MATLAB software to establish daily highest temperature, daily minimum temperature, daily average temperature and humidity of meteorological factors on the regression of power load forecasting model, the specific comparison shows the BP neural network with two kinds of NARX neural network regression prediction results, through the hidden layer of network parameters and the debugging for the appropriate improved BP neural network.