摘要
提出了一种新颖的配电网空间负荷预测方法,采用粗糙集(RS)数据挖掘方法对可能影响小区用地决策的相关属性进行约简,去除冗余属性,克服了以往方法受人为因素影响较大的弱点,得出决定小区用地类型的决策规则,从而得到每个小区适于发展各类负荷的程度,再运用全局最优的土地分配算法来计算各小区内各类负荷的增长,克服了仿真法不能得到整个土地最优分配方案决策的弱点。最后用实例说明了该方法的有效性。
Spatial load forecasting is a process of distributing the total predicted load to all partitioned area, and land-use decision is its key step. In traditional method of spatial load forecasting, land-use decisions mainly depend on the judgement of expert. In this paper a novel method of spatial load forecasting is presented with the help of rough set data mining approach. The shortcomings of traditional methods, which can be affected greatly by human factors, are overcome. Decision-making rules are obtained by which partitioned area's developable degree is described. Furthermore, the whole-optimal land distribution arithmetic is applied to calculate the various increasing load in partitioned area. Shortcomings of simulation, which queues according to the assessment of partitioned area, are avoided because the optimal distribution algorithm can not be obtained by it. Finally, an actual example illustrates the efficiency of the method.
出处
《电工技术学报》
EI
CSCD
北大核心
2005年第5期98-102,共5页
Transactions of China Electrotechnical Society
关键词
空间负荷预测
粗糙集
数据挖掘
分配算法
Spatial load forecasting, rough set, data mining, distribution algorithm
