基于CRS-LMD和SVD的MMC-HVDC线路故障测距方法

直流输电线路故障行波波速不确定、波头提取困难以及噪声干扰等因素制约了直流电网中故障测距技术的应用。为了降低上述因素对定位准确性的影响，提出一种基于局部特征有理样条插值均值分解(LMD based on characteristic rational spline, CRS-LMD)和奇异值分解(singular value decomposition, SVD)的故障测距方法。首先，利用特征尺度选取最优极点系数，结合有理样条插值调节拟合曲线的松紧程度，实现对故障电压行波的局部均值分解。其次，采用奇异值分解对故障行波波头进行准确提取。最后，在PSCAD/EMTDC中搭建了张北±500 kV柔性直流电网的仿真模型，模拟各种故障情况并输出故障数据，利用Matlab对故障数据进行处理并验证定位算法。最后，仿真结果表明，所提故障测距算法在不同故障距离和故障类型下均能实现故障测距，且在叠加噪声和过渡电阻的情况下也能保障较高的精确性。

Fault location of an MMC-HVDC line based on CRS-LMD and SVD

Uncertainty on the fault traveling wave velocity, the difficulty of wave head extraction, the randomness of noise interference and other factors in DC lines restrict the application of fault location technology in a DC power grid. To reduce the influence of the above factors on positioning accuracy, a fault location method based on characteristic rational spline LMD (CRS-LMD) and singular value decomposition (SVD) is proposed. First, the optimal pole coefficient is selected with the use of the feature scale, and tightness of the fitting curve is adjusted by rational spline interpolation. The local mean decomposition of the fault voltage traveling wave is realized. Secondly, SVD is used to extract the fault travelling wave head accurately. Finally, a simulation model of the Zhangbei ±500 kV flexible DC power grid is built in PSCAD/EMTDC to simulate various fault situations and output fault data. Matlab is used to process the fault data and verify the positioning algorithm. The results show that the proposed algorithm can realize fault location with different fault distances and fault types, and guarantee high accuracy under when there is superimposed noise and transition resistance.