基于频域反射系数谱的电缆故障定位与故障类型识别方法研究

Research on cable fault location and fault type identification based on frequency domain reflection coefficient spectrum

频域反射法(FDR)是目前对电缆接地故障诊断较为有效的方法之一。针对输入阻抗谱法识别电缆接地故障在下限频率较高时容易出现误判的问题,本文提出了一种新的基于反射系数谱法的电缆故障定位与故障类型识别方法。首先对接地故障的定位与识别原理进行了分析,基于此对FDR测试频率下限引起的相位误差进行了修正。将布莱克曼自卷积窗的快速傅里叶变换插值算法用于反射系数谱法,同时引入恢复系数对窗函数造成的幅值误差进行修正。建立了不同类型电缆故障仿真模型验证算法的可行性,同时在实验室的40 m同轴电缆与500 m交联聚乙烯电力电缆上对提出算法的准确性进行检验。结果表明:本文所提方法可以准确定位电缆故障,定位误差为0.14%,并有效识别过渡电阻在0~40Z_(0)及开路故障。相较于输入阻抗谱识别法,本研究所提出的反射系数谱识别法更具有普适性,对电缆接地故障定位识别有重要参考意义。

Frequency domain reflectometry(FDR) is one of the most effective methods for cable grounding fault diagnosis. Aiming at the problem the input impedance spectrum method is prone to false identify the cable grounding fault when the lower limit frequency is high, a new cable fault location and fault type identification method based on the reflection coefficient spectrum method is proposed in this paper. Firstly, the principle of fault location and identification is analyzed, based on which the phase error caused by the lower limit of FDR test frequency is corrected. The fast Fourier transform interpolation algorithm of Blackman self-convolution window is applied to the reflection coefficient spectrum method, and the recovery coefficient is introduced to correct the amplitude error caused by the window function. The simulation models of different types of cable faults are established to verify the feasibility of the algorithm, and the accuracy of the proposed algorithm is tested on 40 m coaxial cable and 500 m cross-linked polyethylene(XLPE) power cable in the laboratory. The results show that the proposed method can accurately locate the cable fault, the positioning error is 0.14%, and effectively identify the transition resistance between 0~40Z_(0) and open circuit fault. Compared with the input impedance spectrum identification method, the reflection coefficient spectrum identification method proposed in this paper is more universal, and has important reference significance for cable grounding fault location and identification.