摘要
提出顾及空间异质性的多尺度空间负荷预测模型。提出空间变异系数和尺度的概念,在此基础上提出按照空间变异系数对元胞空间进行不规则区域划分的方法,将得到的分区按照不同的相似度阈值进行区域聚类融合,得到不同尺度下的区域划分,然后将每一尺度下得到的结果进行叠加来预测空间负荷的分布。实例验证表明,所提模型提高了空间负荷预测的准确率。
A multi-scale spatial load forecasting model considering spatial heterogeneity is proposed. The concept of spatial variation coefficient and scale is proposed,based on which,a method is proposed to divide the cellular space into irregular region partitions according to the spatial variation coefficient. The obtained partitions are treated by the regional clustering integration according to different similarity thresholds of spatial forecasting. to get the regional divisions for load. Practical example shows different that the scales,which are superposed to forecast the distribution proposed model improves the accuracy of spatial load
出处
《电力自动化设备》
EI
CSCD
北大核心
2014年第2期91-96,共6页
Electric Power Automation Equipment
基金
国家自然科学基金资助项目(61273144)
北京市自然科学基金资助项目(4122071)~~
关键词
多尺度分析
空间异质性
元胞自动机
C5
0决策树
负荷预测
模型
multi-scale analysis
spatial heterogeneity
element cellular automata
C5.0 decision tree
electric load forecasting
models
