基于ELM与DBSCAN的微电网不良数据检测方法

摘要

Bad data detection can provide reliable data dependence for operation decision-making of micro-grid.Due to the frequency of operation mode switching and difficulty of micro-grid analytical modeling,traditional bad data detection method based on state evaluation has not been applied to micro-grid.This paper utilized extreme learning machine (ELM) to learn the historical data of micro-grid for purpose of extracting the data feature,and detected bad data by DBSCAN clustering algorithm to analyze the feature.A bad data detection method based on ELM and DBSCAN is proposed.Taking advantage of the historical operation data of a four-terminal DC micro-grid prototype,the simulation scenario was designed and the result verified the effectiveness of the proposed method.In addition,this paper contrasted it with several data mining algorithms.It is indicated that ELM + DBSCAN has high superiority on both algorithm performance and detection effects.%不良数据检测可以为微电网运行决策提供可靠的数据依据.由于微电网运行模式切换频繁且系统解析模型难以建立,传统基于状态估计的不良数据检测方法尚未得以应用.文章利用极限学习机(ELM)对微电网历史数据进行学习以提取数据特征;进而利用DBSCAN聚类算法分析特征量以识别不良数据,提出了一种基于ELM和DBSCAN的微电网不良数据检测方法.利用一台四端环形直流微电网样机的历史运行数据构建仿真算例,验证了所提方法的有效性;并与多种常用的数据挖掘算法进行对比,结果表明ELM+DBSCAN在算法性能与检测效果上均具有优越性.