基于边缘计算的电力系统稳态数据压缩方法

Steady-state data compression method for power system based on edge computing

为有效压缩电力系统数据,缩减储存空间与数据传输量,面向电力系统工作过程中生成的大规模稳态数据,以边缘计算为技术支撑,提出一种数据压缩方法实现有效解决。通过建立联合稀疏模型、建立稀疏冗余字典、明确测量矩阵、建立联合重构算法等阶段,结合压缩感知与分布式信源编码,完成边缘计算下电力系统稳态数据融合。利用小波变换算法,按分辨率将得到的数据融合结果分解至各个尺度水平上,得到高频系数与低频系数,经阈值处理高频系数后,采用无损编码技术输出压缩结果。试验阶段中,针对某电网公司的试运行电力系统静态数据展开压缩试验,依据数据压缩空间占比、赋范均方误差以及数据压缩比率3个指标的定量评估结果,检验出所提方法具有较为明显的压缩优势,压缩信号与实际采样信号波形拟合程度较高。

In order to effectively compress power system data,reduce storage space and data transmission volume,and face the large-scale steady-state data generated during the working process of power system,a data compression method was proposed to achieve an effective solution with edge computing as the technical support.Through the establishment of joint sparse model,establishment of sparse redundancy dictionary,clear measurement matrix,establishment of joint reconstruction algorithm and other stages,combined with compressed sensing and distributed source coding,the steady-state data fusion of power system under edge computing was completed.Using the wavelet transform algorithm,the obtained data fusion results were decomposed into various scale levels according to the resolution,and high-frequency coefficients and low-frequency coefficients were obtained.After thresholding the high-frequency coefficients,lossless coding technology was used to output the compression results.In the test phase,a compression test was carried out for the static data of a power grid company′s power system in trial operation.According to the quantitative evaluation results of the three indicators of data compression space ratio,normed mean square error and data compression ratio,it was verified that the proposed method has obvious advantages.The advantage of compression was that the compressed signal had a higher degree of fitting with the actual sampled signal waveform.