摘要
基于时序分析的空间负荷预测方法通常先把预测区域划分成若干元胞,然后对每个元胞进行基于时序的负荷预测,从而实现对整体的空间负荷预测。目前基于时序分析的负荷预测模型有很多,但其中任意单一模型均难以做到对每个元胞预测都是误差最小的,即无法保证对总区域的空间负荷预测结果的准确性。为此提出了一种基于聚类分析理论的空间负荷预测方法,该方法通过分析元胞运用不同负荷预测模型进行预测得出的预测结果和相对误差。实际算例结果表明、该方法正确、有效、可行。
The spatial load forecasting based on time series analysis usually divides the forecasting area into a certain number of structure cells, of which the load forecasting based on time series is then carried out to realize the overall spatial load forecasting. There are lots of load forecasting models at present, any single model, therefore, is difficult to realize the minimum error of every structure cell forecasting, which means that the accuracy of the overall spatial load forecasting cannot be guaranteed. This paper, aiming at this problem, proposes the spatial load forecasting based on cluster analysis, which analyzes the forecasting and relative error of structure cells through different load forecasting models.
出处
《黑龙江电力》
CAS
2011年第6期434-437,440,共5页
Heilongjiang Electric Power
关键词
空间负荷预测
聚类分析理论
元胞
spatial load forecasting
cluster analysis
structure cell
