数据挖掘下电网调度信号异常数据提取方法  

Abnormal data extraction method of power grid dispatching signal under data mining

作　　者：刘峰[1] 朱颉[1] 张凯[1] 冯晗[1] 张伟[1] 周庆捷 LIU Feng;ZHU Jie;ZHANG Kai;FENG Han;ZHANG Wei;ZHOU Qing-jie(State Grid Zhengzhou Power Supply Company, Zhengzhou 450006,China;School of Electrical and Electronic Engineering, North China Electric Power University, Beijing 102206,China)

机构地区：[1]国网郑州供电公司,河南郑州450006 [2]华北电力大学电气与电子工程学院,北京102206

出　　处：《湘潭大学学报（自然科学版）》2021年第6期74-80,共7页Journal of Xiangtan University(Natural Science Edition)

基　　金：国网河南省电力公司郑州供电公司电力规划科技项目(zhbr-ht-2021-0829)。

摘　　要：目前方法获取电网调度信号特征过程中,针对在异常数据提取时提取精度不高、数据覆盖率欠高、提取效果差的问题,提出数据挖掘下电网调度信号异常数据提取方法.首先利用数据挖掘中的ICA算法对电网调度信号去噪处理;再利用Mallat算法将电网调度信号进行小波分解,获取电网调度信号的信号特征;最后通过对电网调度信号数据中异常值计算,获取异常数据检测阈值,完成电网调度信号异常数据提取.实验结果表明,运用该方法提取信号中的异常数据时,数据提取的精度高、数据覆盖性能好、提取的效果好.At present,in the process of obtaining the characteristics of power grid dispatching signal,aiming at the problems of low extraction accuracy,low data coverage and poor extraction effect in abnormal data extraction,an abnormal data extraction method of power grid dispatching signal under data mining is proposed.Firstly,the ICA algorithm in data mining is used to denoise the power grid dispatching signal;Then the Mallat algorithm is used to decompose the power grid dispatching signal by wavelet to obtain the signal characteristics of the power grid dispatching signal;Finally,the abnormal data detection threshold is obtained by calculating the abnormal value in the power grid dispatching signal data,and the abnormal data extraction of power grid dispatching signal is completed.The experimental results show that when using this method to extract the abnormal data in the signal,the data extraction accuracy is high,the data coverage performance is good,and the extraction effect is good.

关 键 词：数据挖掘下 电网调度信号 异常数据 提取方法 MALLAT算法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]