基于相似日的神经网络短期负荷预测方法

人工神经网络是模拟人脑神经元结构、特性和大脑认知功能而构成的新型信号、信号处理系统。本文针对电力负荷短期预测问题，提出了一种基于相似日的神经网络预测方法，采用反向传播算法，考虑气象因素对负荷的影响，提高了学习效能，具有较好的预测精度。本方法很适合在短期负荷预测中使用，预测结果验证了上述结论。

SHORT-TERM LOAD FORECASTING USING A NEURAL NETWORKBASED ON SIMILAR HISTORICAL DAY DATA

Neural network,which simulates the frame,the character and the cognizing ability of nerve element cerebrum,is a new system that can handle many kinds of information semaphore.Aiming at short term load forecasting,this paper introduced an ANN based on similar historical day data.This method adopted BP arithmetic and considered weather factor,which could improve the learning property and perform better forecasting accuracy.The method is generally suitable for short term load forecasting.The forecasting result proved the above conclusion.