FNN-Based Ground Fault Identification Method for 10kV Distribution Networks

With the development of power systems, the distribution network, as an important link in power supply, has become increasingly crucial in terms of stability and security. To further improve the performance of ground fault identification, this paper proposes a method for identifying ground faults in 10kV distribution networks based on the FNN model. By combining the characteristics of partial fault conditions in distribution networks and based on actual distribution network parameters, a 10kV distribution network model is built in PSCAD. Various ground fault sequence simulations are conducted by introducing a fault module to obtain experimental data. Using the obtained sample data, a study on ground fault type identification in distribution networks based on the FNN model is constructed. The experimental results show that the loss function value ranges from 0.3 to 0.4, with an overall classification accuracy rate exceeding 90%. Furthermore, compared to the performance of other models, this method significantly outperforms other models in terms of classification results for this type of task.