摘要
数据挖掘技术能够从大量数据中发现潜在知识 ,软计算是创建智能系统的有效方法 ,本文将两者结合 ,完成电力预测过程的两个主要任务 :负荷坏数据处理和多因素负荷预测模型的建立。通过对 Kohonen网聚类挖掘和 BP网分类挖掘的效果分析 ,设计由这两种网络组合而成的神经网络模型 ,完成坏数据辨识和调整的任务 ;以模糊推理系统为基础构建多因素负荷预测模型 ,本文采用 CART分类挖掘技术解决模糊结构辨识中的两个难点问题 :输入空间划分和输入变量选择 ,在此基础上设计 ANFIS网络进行参数辨识。良好的实例分析效果说明 ,数据挖掘思想和软计算方法相结合 。
Data mining distills co nn otative knowledge and information from abundant data, while soft computing is an effective method to establish intelligent computing systems. This dissertation combines the two methods to accomplish two main tasks: outliers processing in lo ad data and multi-factor STLF (short term load forecasting) system modeling. By analysising the effects of Kohonen network clustering and BP network classificat ion, the dissertation designs an outlier identification model comprising these t wo kinds of neural network and implements the tasks of bad data identifications and adjustments. A STLF system is modeled based on ANFIS (Adaptive Neural-Fuzzy Inference System). CART can solved two difficulties in the fuzzy structure ident ification: the variables selection and the input space division. And then an ada ptive network is constructed to identify the parameters. Good test results using actual data demonstrate that combining of date mining and soft computing is a g ood idea and method in STLF.
出处
《电力系统及其自动化学报》
CSCD
2003年第1期1-4,94,共5页
Proceedings of the CSU-EPSA
关键词
负荷预测
模颧推理
数据挖掘
软计算
电力系统
分类
聚类
Load forecasting, Fuzzy inference, Data mining, Classification and clustering, Soft computing
