摘要
电力系统电力消费量受诸多因素的影响,为了提高其预测的精度,得到更好的预测结果,首先分析了异常数据产生的原因以及其对预测结果的影响,提出了基于M-估计的稳健回归预测算法。在该预测算法中首先选择恰当的目标函数和权重函数,接着利用泰勒展开对参数进行迭代估计,得到稳健的预测值,最后进行算例分析。算例数值表明该算法同传统线性回归方法相比,具有较好的抗干扰性,预测结果能更好地吻合实际数据和未来的趋势。
Electric power consumption is influenced by many factors. To improve its forecasting precision and obtain better results, the cause and effect of anomalous data are analyzed, and a robust regression algorithm based on M- esti- mation is given. In the algorithm, proper objective function and weight functions are chosen, and then iterative estima- tion using Taylor expansion is made to obtain robust forecasting results. In the end, experiment results show that the proposed method performs better both in robustness and in accuracy than the conventional linear regression method.
出处
《电力系统及其自动化学报》
CSCD
北大核心
2013年第5期22-25,共4页
Proceedings of the CSU-EPSA
基金
国家自然科学基金项目(60874113)
关键词
最小二乘法
M-估计
稳健回归
预测
电力消费
optimal least square
M- estimation
robust regression
forecasting
electric power consumption
