基于Nuttall自卷积窗的10 kV配电电缆中间接头定位改进方法

为解决现阶段采用离散傅里叶变换（discrete Fourier transform，DFT）分析电缆首端输入阻抗谱时存在频谱泄露和栅栏效应的问题，提出一种基于Nuttall自卷积窗的10kV配电电缆中间接头定位改进方法。首先，应用含有中间接头的电力电缆传输线模型、电缆输入阻抗谱、中间接头定位原理、加窗傅里叶变换（windowed Fourier transform，WFT）简要说明本文所提方法原理。接着，选取窗函数（Hamming窗、Blackman窗、四项三阶Nuttall窗、二阶Nuttall自卷积窗），详细分析窗函数性能。然后，通过对中间接头的定位仿真，分析比较4种窗函数的WFT算法对接头的识别定位效果。最后，对实验室500 m长的10 kV交联聚乙烯电缆（cross-linked polyethylene，XLPE）进行测试来检验所提方法的测试效果。仿真和实验结果表明：相比于Hamming窗、Blackman窗、四项三阶Nuttall窗，二阶Nuttall自卷积窗抑制频谱泄露和栅栏效应的效果最佳，能更高灵敏度地识别和定位中间接头，且定位误差小于0.024%。

An Improved Location Method of 10kV Cable Joints Based on Nuttall Self-Convolution Window

In order to solve the problem of spectrum leakage and fence effect when using discrete Fourier transform(DFT)to analyze the input impedance spectrum of header point of cable,this paper presents an improved method of the joints location for the 10kV distribution cable based on Nuttall self-convolution window.First,the principle of the proposed method was briefly explained in this paper,including transmission line model for the power cable with an intermediate joint,input impedance spectrum of the cable,the locating principle of joints,and the windowed Fourier transform(WFT).Next,the window functions,i.e.Hamming window,Blackman window,four-iterm and third-order of Nuttall window,second-order of Nuttall self-convolution window,were selected to analyze their characteristics in detail.Then,through the simulation,the effects of the WFT algorithm with four window functions were analyzed and compared for the middle joint location.Finally,a 500 m long 10 kV cross-linked polyethylene(XLPE)cable was tested to verify the test results in the laboratory.The results indicated that compared with Hamming window,Blackman window and the four-iterm and third-order Nuttall window,the second-order of Nuttall self-convolution window had the best effect of suppressing spectrum leakage and fence effect.Besides,the proposed location method can identify and locate joints with higher sensitivity.The positioning error was less than 0.024%.