摘要
短期负荷预测是指预测未来24 h内的电力负荷需求,这是一项非常重要的工作。目前,负荷预测的实用计算方法有很多,线性回归法、时间序列法、人工神经网络法等等,但是,这些算法的预测精度欠佳。现根据相似日负荷的相似性,提出利用聚类分析法来进行短期负荷预测,为此,介绍聚类分析的步骤、方法及验证算例。实际运行结果表明:利用聚类分析法进行负荷短期预测,短期负荷预测的精度大大提高。
Short-term load forecasting is to forecast the power load demand in the future 24 hours. At present, there are many practical algorithms for load forecasting, such as the methods of linear regression, time series and artificial neural network. However, the accuracy of these methods is unsatisfactory. According to the likeness of the load in resembling days, the use of cluster analysis method is presentcd for short-term load forecasting. The approach and the verifying example of cluster analysis are described. Practical operation results show that the cluster analysis method can considerably improve the accuracy of short-term load forecasting.
出处
《广东电力》
2006年第1期18-21,共4页
Guangdong Electric Power
关键词
短期负荷预测
电力负荷特性
聚粪分析法
人工神经网络法
short-term load forecasting
power load characteristics
cluster analysis method
method of artificial neural network