摘要
提出考虑供电量组分多层级划分及外部因素影响,利用关联度寻优方法构造最优GM(1,N)电量预测模型。根据供电地区的行业用电分类,对总供电量的组分进行分层级划分和重要性分析;计算各重要组分及外部影响因素与供电量的关联度,并依据关联度大小对各影响因素进行排序,再通过建立不同N下的GM(1,N)模型,根据预测精度确定最优GM(1,N)模型。采用该模型对广东电网中山供电局的供电量数据进行预测分析,证明了该模型的预测结果具有较高的准确性。
This paper proposed to use the correlation optimization method to construct the optimal GM （1, N） electric quantity predictionmodel considering the power supply components multilevel division and external influencing factors. According to the industry powerutilization classification of power supply area, this paper carried out the power supply components multilevel division, analyzed the importanceof the power supply components and calculated the correlation between each important component, together with external influencingfactors and the power supply components. Each influencing factor was sorted based on the correlation and the GM （1, N） model ofdifferent N was established to determine the optimal one according to the prediction accuracy. The actual power supply data of Zhongshanpower supply bureau of Guangdong power grid verifies the high accuracy of this model’s forecasting algorithm.
出处
《电工电气》
2018年第1期27-31,37,共6页
Electrotechnics Electric
关键词
季度电量预测
GM(1
N)模型
行业用电分类
外部影响因素
quarter power supply forecasting
GM （1, N） model
industry power utilization classification
external influencing factor
