基于量测大数据和数学形态学的配电网故障检测及定位方法研究

提出基于量测大数据和数学形态学的配电网故障检测及定位方法，该方法基于多点同步测量的采集分析架构可使所有数据在同一个时间剖面。首先，通过同步量测采集到的行波信号和工频信号，结合配电网拓扑结构，构建全网大数据矩阵；然后利用Trace检测、圆环定律等方法判别波形信号奇异点，快速判别是否存在故障及不同类型故障的波形畸变规律，从而达到快速故障检测的目的；在检测到故障后，采用基于现代D型行波故障定位方法对故障点进行高精度定位。最后，在PSCAD 80节点配电网模型中对该方法进行了仿真验证，验证结果说明本文提出的方法具有不受线路长度和波速影响，精确度较高的特点。

Distribution Network Fault Location and Detection Method Based on Measurement Big Data and Mathematical Morphology

This paper proposes the fault detection and location method for distribution network based on measurement big data and mathematical morphology， which is based on the acquisition and analysis architecture of multi-point synchronous measurement to make sure all the data is in the same time profile. Firstly，the traveling wave signal and the power frequency signal collected by synchronous measurement is combined with the topology structure of the distribution network to construct a large data matrix of the whole network. Then the faults and waveform distortions of different types of faults are quickly determined by using Trace detection and ring law to discriminate the singularity of the waveform signal，achieving the fast fault detection. After a fault is detected， the high-precision positioning of the fault points is done using modern D-type traveling wave based fault location method. Finally， the method is verified with the PSCAD 80-node distribution network model. The results show that the proposed method is of high accuracy and not be affected by line length and wave speed.