基于MRSVD-VMD的多分支配电网故障定位

Fault Location of Multi-branch Distribution Network Based on MRSVD-VMD

针对配电网行波定位中受噪声干扰导致波头标定困难和传统定位方法不适用于多分支配网结构的问题,提出基于多分辨率奇异值分解-变分模态分解MRSVD-VMD(multi-resolution singular value decomposition-variational mode decomposition)的自适应波头标定算法和不受行波波速影响的T域定位算法。利用MRSVD和VMD分解故障行波,根据峭度值和峭熵比筛选有效分量,然后通过对称差分能量算子SDEO(symmetrical differencing energy operator)实现波头标定;最后利用行波到达时间筛选故障T域,实现故障点的区段定位和精确测距。仿真结果表明,MRSVD-VMD行波波头标定方法在不同噪声下能有效标定波头,T域定位算法排除波速影响,能实现多分支配电网故障的精确定位。

Aimed at the difficulty in wave head calibration due to noise interference when locating a travelling wave in a distribution network and the problem that the traditional location methods are not suitable for a multi-branch distribution network,an adaptive travelling wave head calibration algorithm based on multi-resolution singular value decomposition-variational mode decomposition(MRSVD-VMD)and a T-domain location method which is not affected by the travelling wave speed are proposed.MRSVD and VMD are used to decompose the fault travelling wave,and the effective component is selected according to the value of kurtosis and the ratio of kurtosis to entropy.Then,a symmetric differencing energy operator(SDEO)is used to calibrate the wave head.Finally,the traveling wave arrival time is used to screen the fault T-domain,thus realizing the location of the segment containing the fault point as well as the accurate distance measurement.Simulation results show that the MRSVD-VMD algorithm can effectively calibrate the wave head under different noises,and the T-domain location algorithm can avoid the effects of wave speed,thereby realizing the accurate location of fault in the multi-branch distribution network.