基于回归BP神经网络的短期负荷预测

负荷的形成受多方面因素的影响，在建立短期负荷预测模型时，需要综合考虑多种因素。同时，负荷是一种时间序列信号，目前的数据能够对以后的数据产生重要的影响，所以文章采用回归BP神经网络模型应用于短期负荷预测。实例计算表明，该方法有效，预测精度比常规方法高，收敛性好，运算速度快。

Short-term Load Forecasting Based on Recurrent BP Neural Networks

The load forming is affected by various factors. In this case, when short-term load forecasting model is being built, many factors should be taken into consideration. Meanwhile, load is sequence signal of time. Therefore, the data at present would have effects on the data in the future. The article adopts the model of recurrent BP neural networks applying in short-term load forecasting. Evidenced with the calculation, this method is effective and more precise than common methods, and features with good convergence and fast operation speed.