基于RBF神经网络和专家系统的短期负荷预测方法

深入研究了天气和特殊事件对电力负荷的影响，建立了结合径向基（RBF）神经网络和专家系统来进行短期负荷预测的模型。利用RBF神经网络的非线性逼近能力预测出日负荷曲线，然后利用专家系统根据天气因素或特殊事件对负荷曲线进行修正，使其在天气突变等情况下也能达到较高的预测精度。利用该模型编制的实用化软件在西北电网的多个电力局投入实际应用，结果表明：该方法用BP神经网络相比，具有较高的预测精度，同时具有较强的实用性。

Short-Term Load Forecasting Using Radial Basis Function Networks and Expert System

Investigating the effect of weather factors and special events on electric power load, a load forecasting model based on RBF(Radial Basis Function) neural networks and expert system is established, and an effective algorithm is also designed. First the load curve for one day was approximated by RBF net using its nonlinear convergence. Then expert system was used to make this model work under special disturbance. A practical software package has been formed and applied to some Power Nets of Northwest China Power System, which improves the precision of short term load forecasting. The effectiveness of the model has been verified by actual operation.