Page 40 - 电力与能源2024年第三期

第 45 卷 第 3 期
电力与能源
314 2024 年 6 月

DOI：10.11973/dlyny202403006

基于改进变分模态分解的电网企业

监测数据滤波研究

史渊源，万 鹏

（国网宁夏电力有限公司，宁夏 银川 750001）

摘 要：电网企业监测数据在采集过程中，常会受到设备误差、环境因素等多种噪声和干扰的影响，这些干扰
因素会使得数据呈现出复杂结构和非平稳性的特征。传统的数据滤波方法在处理这类信号时，对于噪声干扰
的容忍度相对较低，会过度平滑或削弱信号中的有用信息，导致数据质量无法得到显著提升。为此，设计了一
种基于改进变分模态分解（IVMD）的电网企业监测数据滤波方法。在采集电网实时监测数据后，对采集的这
些数据进行小波变换。通过对信号进行不同尺度的分解，实现对电网企业监测数据原始信号的重构。设置阈
值函数对小波变换后的数据进行处理，去除噪声并保留有用信号，提高信号的纯净度和信噪比。为了进一步
提升数据的准确性和可靠性，利用 IVMD 方法对于去噪后的数据进行进一步的处理，以滤除剩余的噪声或干
扰成分。试验结果表明，所设计的方法滤波效果显著，其失真比最小值达到了 0.028 dB，这表明通过该方法能
够得到更加纯净和准确的电网企业监测数据。
关键词：改进变分模态分解；电网企业；监测数据；滤波
作者简介：史渊源（1982—），男，高级工程师，从事电力企业大数据分析、运营监测工作。
中图分类号：P208 文献标志码：A 文章编号：2095-1256（2024）03-0314-06
Filtering of Power Grid Enterprise Monitoring Data Based on Improved Variational

Mode Decomposition

SHI Yuanyuan，WAN Peng

（State Grid Ningxia Electric Power Co.，Ltd.，Yinchuan 750001，Ningxia Hui Autonomous Region，China）

Abstract：During the collection process of power grid enterprise monitoring data, various noises and interferences
such as equipment errors and environmental factors often affect the data. These interfering factors result in the
data exhibiting complex structures and non-stationary characteristics. Traditional data filtering methods, when
dealing with such signals, have relatively low tolerance for noise interference, leading to excessive smoothing or
weakening of useful information in the signals, thereby failing to significantly improve the data quality. Therefore,
a filtering method for power grid enterprise monitoring data based on improved variational mode decomposition is
designed. After collecting real-time monitoring data from the power grid, wavelet transform is applied to these
collected data. By decomposing the signals into different scales, the original signals of power grid enterprise moni⁃
toring data are reconstructed. A threshold function is employed to process the data after wavelet transform, remov⁃
ing noise while retaining useful signals, thereby enhancing the purity and signal-to-noise ratio of the signals. To
further improve the accuracy and reliability of the data, the improved variational mode decomposition (IVMD)
method is utilized for further processing the denoised data to filter out residual noise or interference components,
thus enhancing the accuracy and reliability of the signals. Experimental results show that the filtering effect of the
proposed method is significant, with the minimum distortion ratio reaching 0.028 dB, indicating that the method
can obtain cleaner and more accurate power grid enterprise monitoring data.

Key words：improved variational mode decomposition，power grid enterprise，monitoring data，filtering

随着电力行业的迅速发展，电网企业对电网 运行安全和稳定性的要求也在日益提高。电网企