Study On The Fault Line Detection Using Information Gain Degree And Support Vector Machines For Small Current Grounding System

Fault line detection for small current grounding system has always been a difficult point in the field of relay protection. Because single line detection method has many limits and single-phase earth fault is multiplicity and changeable, the accuracy of fault line detection is very low. In order to improve the accuracy of fault line detection, the combination of many single line detection methods using information fusion technology is studied. In view of the above backgrounds, this paper carries on the depth research to this problem.Firstly, this paper analyzed the single-phase grounding faults mechanisms of the small current neutral grounding system, especially the steady and transient fault characteristics. With the simulation model established by MATLAB, line detection methods are verified and their adaptability is analyzed.Secondly, the fault feature is extracted from the zero-sequence component by wavelet method and fast fourier transform, and fault measure function is established by means of the fault feature. And then the weigh of single line detection methods is given by information gain degree. When breaking down, the line fault is judged by the result of single line detection multiplied in the form of the weigh. The simulation shows that the accuracy of fault line detection is improved.In order to further improve the correct rate of fault line detection, support vector machine whose parameters are optimized by particle swarm algorithm is applied to global diagnosis by sampled data of every method respectively, and weights of line fault are given using information gain degree. The results showed that the method that applied to fault line detection is accurate and sensitive through simulation system.