应用人工神经网络进行短期负荷预测

SHORT-TERM LOAD FORECASTING USiNG ARTIFICIAL NEURAL NETWORK

本文提出了一种应用人工神经网络进行电力系统短期负荷预测的方法。负荷按照每周各日进行分类，共七种模式，学习样本选取每周中的相同类型日。为了提高预测精度，对原始数据中的伪数据进行清除，对于那些可以预料到的随机干扰，应用专家系统原理予以处理。通过对银川供电局负荷的实际预测，表明本文所提供方法可以实际应用。

In this paper we propose a method for short-term load forecasting using artificial neural network.Therer are seven didstinct load patterhs:from monday to sunday. Learning sample come from the previous load data witht arme pattern.We eliminate durn my data in original data and dispel random disturbance using principles of expert system. The results of Yinchuan load forecasting system showed effectiveness of the proposed method.