A fault line selection method for small current grounding system based on big data

摘要

The fault mechanism of small current grounding system for single-line-to-ground (SLG) faults and the functions of big data techniques are discussed. A new fault line selecting method for SLG faults based on big data theory is proposed. The method resolves the problem of detecting fault lines by making use of a mass of data from the grid which consists of both electrical and non-electrical quantities. Some unstructured data is preprocessed to feature information of faults, and data mining techniques are used to learn the complex nonlinear mapping relation between data and fault lines. Whenever a new grounding fault occurs, the mapping relation can be used to select the fault line. The small current grounding system is modeled by simulation software PSCAD and data is analyzed by a software package MATLAB and SPSS Modeler, and the result shows that this method works effectively and improves the accuracy of fault line selection.