摘要
电力系统负荷预测中，序列量一般采用多种方法进行预测，得到多种结果。基于这些结果，提出了以各时段残差平方和最小为目标函数的负荷预测的综合模型，并根据模型特点，给出了简捷直观的求解方法。进一步分析了模型的物理意义，提出了一个近似的综合预测模型，从而在保证一定精度的前提下减少计算量。研究了该方法在电力系统负荷预测中的应用，分析结果表明，这两种综合预测模型的拟合精度均明显优于各种单一模型，并能得到更好的预测效果。
Multiple candidate methods are usually used in serial forecasting so that various results can be got. In order to minimize squares sum of the residues, an integrated forecasting model based on these results is proposed. A simple solution method of the model is also presented according to its feature. An approximate model is further got to speed up its calculation. The case study shows that both these integrated models are better than each single one and more accurate results can be got.
出处
《清华大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
1999年第1期8-11,共4页
Journal of Tsinghua University(Science and Technology)
关键词
电力系统
负荷预测
综合模型
非线性规划
power system load forecasting
integrated model
squares sum of residues
nonlinear programming
