Abstract: Electrical insulator leakage current is an important parameter to characterize the operating status of insulator. It needs large amounts of data to analyze the characteristics in high frequency, which increases data transmission burden. Data compression method can effectively reduce the amount of data and decrease requirements of transmission bandwidth. According to the existing data compression method and the characteristics of leakage current data, it proposes a method. It uses EMD（Empirical Mode Decomposition） to decompose the data, then uses the fractal theory to compress and recover the leakage current data. The results show that it can get higher compression ratio and smaller reconstruction error than the present data compression method and also meets the validity and the superiority.

Key words: insulator leakage current, data compression, fractal theory, Empirical Mode Decomposition（EMD）

摘要： 电力绝缘子泄漏电流是表征绝缘子运行状态的重要数据，但为了分析其高频特性，采样数据量需要很大，加重了数据传输负担。数据压缩方法能够有效减小数据量，降低传输带宽要求。根据现有数据压缩方法，结合泄漏电流数据特征，提出了利用EMD（Empirical Mode Decomposition）对数据进行分解，并用分形理论对泄漏电流数据进行压缩和恢复的方法。实验结果表明，它比现有数据压缩方法能够得到更高的压缩效率和更小的重构误差，证明了该方法的有效性和优越性。

关键词: 绝缘子泄露电流, 数据压缩, 分形理论, 经验模态分解（EMD）