摘要
对配电系统短期负荷预测的周期自回归模型和配电负荷的周期性进行了研究。采用相关分析法对配电负荷的周期特性作了深入地分析,研究结果表明配电负荷的日周期性比周周期性更明显;基于配电负荷的时刻相关性分析,挑选出对预测结果起决定性作用的特征输入量,据此提出了改进的配电负荷日周期PAR预测模型。实例研究表明,该模型较常规PAR预测模型的预测速度更快、精度更高。
The periodical auto-regression（PAR）models for short term load forecasting in distribution electric systems and the periodicity of load cycle are investigated.The result of the linear correlation analysis of distribution load curves shows that daily periodicity is more obvious than that of weekly.An improved daily period PAR load forecasting model is proposed accordingly.Based on the correlation analysis of the hourly period of distribution load,main determinants affecting load forecasting results are chosen as the characteristic input variables for the model.The forecast results of the example show that this model is of more distinct performance improvement than that of the normal PAR forecasting models.
出处
《电力系统保护与控制》
EI
CSCD
北大核心
2010年第14期128-133,共6页
Power System Protection and Control
基金
'十一五'国家科技支撑计划(2006BAJ04B06)
关键词
短期负荷预测
周期自回归模型
线性相关性分析
配电负荷
特征输入量
short-term forecasting
periodical auto-regression model
linear correlation analysis
distribution load
characteristic inputs
