基于分形测度的电压暂降持续时间检测

Voltage Sag Duration Detecting Method Based on Fractal Number Measurement

基于分形技术,提出了一种新型的电压暂降持续时间检测方法。该方法利用分形数计算信号波形的平滑度,识别暂降起止时刻的高频暂态振荡,然后将实时计算的分形数与阈值进行比较,得到起止时间的定位进而求得持续时间。基于PSCAD/EMTDC软件进行故障暂态仿真,在有无谐波畸变的情况下分别进行测试,证明该方法有效可行。同时进一步将此方法与小波变换及有效值法分析比较,表明其具有较好的实时性和准确性。

Based on fractal technology,a novel voltage sag duration detecting method is proposed. This method calculates the fractal number of signal waveform as an evenness index which is used to detect the high frequency transient oscillation arising at the start and stop time of voltage sag. Then the fractal number is refined through threshold comparison method to accurately reflect the signal dynamics and find the duration of sag event. A fault model was simulated using PSCAD/EMTDC software. The sag waveforms with and without harmonics were detected respectively to verify the feasibility of the proposed method. This method is further compared with wavelet transform and root mean square（RMS）method. The results show its good performance on rapidity and accuracy.