基于神经网络的电力系统短期负荷预测

介绍了短期负荷的特点，深入分析了温度、降雨量、时间等因素对负荷的影响。应用BP神经网络，建立了充分考虑各种因素的日最高负荷预测和日平均负荷预测模型，最后通过贵阳城南分局历史负荷进行验证，表明本论文所建立的神经网络预测模型充分考虑了各种负荷因素影响，预测精度良好，具有较好的非线性映射能力，有进一步开发应用于实际预测的良好前景。

Short- term load forecasting in power system based on neural network

This paper introduces the features of short - term loading, and then analyzes the impact of tem- perature, rainfall, and time on leading. By applying BP neural network, models of daily peak and aver- age load forecasting are established based on all kinds of factors. Finally, the models are verified by the historical loading in Guiyang Power Supply Bureau. The verification shows that all kinds of influential fac- tor are considered in the neural network forecasting model established by this paper, this model have good preciseness in forecasting, and are characterized in good non -linear mapping ability. Thus these models have great potential in application to reality forecasting.