基于时域波形特征的输电线雷击识别

Identification of lightning strikes on transmission lines based on the characteristics of time-domain waveforms

雷击输电线路产生大量暂态分量影响行波测距装置精度,实现雷击干扰与故障的识别对提高行波测量装置准确性、减少检修工作量具有重要意义。为全局应用故障信息避免时频变换产生信号干扰,本文引用近似熵算法对雷击扰动与短路故障暂态电流附加线模分量进行时域分析发现,在不同故障相角与噪声干扰下,两者熵值差异仍接近一个数量级,但该方法不能识别非故障性雷击与雷击故障,故运用信号处理常用积分算法对雷击扰动行波做进一步处理发现,时间轴上下侧积分值的标准方差差异明显,综合两法实现非故障性雷击、雷击故障与短路故障电流暂态行波的区分。大量PSCAD仿真实验证明本方法不受故障位置、故障条件与噪声干扰的影响,具有很强的适用性。

Transient signals caused by lightning strikes on transmission lines may interfere with the distance measuring devices relying on traveling waves. Therefore,it is significant to identify lightning interference and faults for improving the accuracy of the measuring devices and reducing the maintenance workload. In order to make use of the global fault message and avoid the signal interference generated by time-frequency transformation,this paper referred to the approximate entropy algorithm to make a time-domain study on the line-mode components of lightning interference and short-circuit fault transient current and found that under different phase angles and noise interference of different faults,both entropies neatly differed in an order of magnitude. But this method could not distinguish between non-fault lightning and lightning faults. Therefore,this paper took advantage of the integral algorithm which was widely used in signal processing to further analyze the lightning disturbance waves and found that the standard deviation differences of the integral values on the upper and lower timeline were obviously distinctand that the identification of non-fault lightning,lightning faults and short-circuit faults could be realized by integrating these two methods. Numerous PSCAD simulation tests have demonstrated that the proposed method is valid and applicable regardless of the fault location,fault condition and noise interference.