Research On Method Of Fault Line Detection In Small Current Grounding Based On Fuzzy Neural Network

The small current grounding mode is widely used in the low-voltage distribution network at home and abroad. The single-phase earth fault occurs in the small current grounding system mostly. Although in a short period of time it does not affect the normal operation of power grid, but in order to prevent the expansion of the fault, we must find the fault line as soon as possible. Due to the complexity of power distribution network structure, and the limitations of each single line detection method, the problem of fault line detection has not been solved. With the development of the artificial intelligence, the line detection method based on multiple criterion intelligent fusion is applied. The fuzzy neural network to incorporate each line detection method effectively, But there are some problems such as network model construction, being easily trapped in local extreme value and so on. Considering the genetic algorithm has strong global optimization ability, combination with the fuzzy neural network was applied to the fault line detection, which can improve the reliability of line detection.Based on the detailed analysis of the transient steady state characteristics of single-phase earth fault in small current grounding system, with the combination of the existing advantages and disadvantages of the fault line detection method, select the different basis criteria for different grounding method, and the use of information fusion and artificial intelligence technology, a fault line detection method based on improved genetic algorithm for small current grounding fault of fuzzy neural network is proposed in this paper.In this paper, the model of small current grounding system is built by using Matlab/simlink simulation platform. it also verifies the effectiveness and limitations of under different grounding method to select the different basis criterion. The effects of different fault conditions on the transient zero sequence current are analyzed. In view of a small current grounding system of different grounding method, Respectively using improved genetic algorithm to optimize fuzzy neural network simulation, and comparing with the single traditional fuzzy neural network and the optimized fuzzy neural network based on traditional genetic algorithm on the training effect and the ability to select, the paper proves the quickness, accuracy and reliability of the proposed algorithm. At the same time, it also verifies the feasibility of the application of the proposed algorithm in small current grounding fault line with a large number of distributed generation grid.