华东师范大学学报(自然科学版) ›› 2020, Vol. 2020 ›› Issue (4): 156-163.doi: 10.3969/j.issn.1000-5641.201921016

• 计算机科学 • 上一篇    下一篇

基于灰度模型的电能量异常数据修复研究

黄福兴1,2, 周广山1,2, 郑宽昀1,2, 冯泽佳3, 袁培森3   

  1. 1. 南瑞集团有限公司(国网电力科学研究院有限公司), 南京 211106;
    2. 国电南瑞科技股份有限公司,南京 211106;
    3. 南京农业大学 信息科学技术学院, 南京 210095
  • 收稿日期:2019-08-26 发布日期:2020-07-20
  • 通讯作者: 袁培森,男,博士,讲师,研究方向为大数据处理与分析.E-mail:peiseny@163.com E-mail:peiseny@163.com
  • 作者简介:黄福兴, 男, 高级工程师, 研究方向为电能量计量、用电信息采集和智能量测体系. E-mail: huangfuxing@sgepri.sgcc.com.cn

Research on repairing anomalous electrical energy data based on the Grey Model

HUANG Fuxing1,2, ZHOU Guangshan1,2, ZHENG Kuanyun1,2, FENG Zejia3, YUAN Peisen3   

  1. 1. NARI Group Corporation (State Grid Electric Power Research Institute), Nanjing 211106, China;
    2. NARI Technology Co., Ltd, Nanjing 211106, China;
    3. College of Information Science and Technology, Nanjing Agricultural University, Nanjing 210095, China
  • Received:2019-08-26 Published:2020-07-20

摘要: 提出了一种基于灰度模型的电能量异常数据修复方法, 以经过识别的正常历史电能量数据作为输入变量, 以异常点所处的时间节点电能量数据作为输出变量, 经过一次累加, 级比检验, 求解预测方程得到预测值, 动态地对电能量数据进行迭代预测, 最终对预测值进行精度检验, 预测的平均相对残差为2.182%, 根据结果对原始数据进行修改, 从而达到修复电能量异常数据的目的. 以某区域实际电能量数据进行模型预测修复, 并对结果以及误差进行分析, 验证了该方法的可行性.

关键词: 电能量, 异常数据, 灰度模型, 数据修复

Abstract: The traditional technique of repairing anomalous electrical energy data requires large amounts of data, has a high operational cost, and results in poor timeliness by using interpolation and other statistical methods; hence, the accuracy and efficiency of repairing results are limited. In this paper, a method for repairing anomalous electrical energy data based on the Grey Model is proposed. The normal historical electrical energy data is taken as an input variable, and the time node electrical energy data at which the abnormal point is located is taken as the output variable. The ratio test and the prediction equation are used to obtain the predicted value. The electrical energy data is iteratively predicted. Finally, the accuracy of the predicted value is tested. The average relative residual of the prediction was found to be 2.182%. The original data is then modified according to the result so as to repair the electrical energy anomaly data. The model prediction and repair are carried out with the actual electrical energy data of a certain area, and the results and errors are analyzed. The feasibility of the method is subsequently verified.

Key words: electrical energy, anomaly data, Grey Model, data repair

中图分类号: