摘要
针对混沌一阶局域预测模型中欧氏距离不能反映相空间中相点间的相关性大小,提出一种改进加权一阶局域预测模型。该模型将相点间的关联度和欧氏距离以加权值形式作用于一阶线性回归方程,综合考虑了最邻近点对预测中心点的邻近效应以及最邻近点与预测中心点的相关性,克服了常规一阶局域预测法存在的不足。通过实际算例分析显示,本文方法较改进前有较好的适应能力和预测精度。
The correlation between nearest neighboring points and prediction point is not reflected by the Euclid distance in the chaotic local linear forecasting method; an improving adding-weight one-rank local forecasting model is presented in this paper. Euclid distance and degree of incidence between nearest neighboring points and prediction point in the model are considered as an adding-weighting value and applied to one-rank local linear forecasting model. It synthetically takes into account the effect of distance and degree of incidence between nearest neighboring points and prediction point in order to overcome the defect of local linear model. The results show the model can work more effectively.
出处
《电测与仪表》
北大核心
2006年第5期5-8,共4页
Electrical Measurement & Instrumentation
关键词
短期负荷预测
混沌时间序列
欧氏距离
加权一阶局域法
关联度
short-term load forecasting
chaotic time series
Euclid distance
adding weight one-rank local region method
degree of incidence
