Power Grid Fault Diagnosis Based on Immune RBF Neural Network

Radial basis function (RBF) has achieved good application in power grid fault diagnosis because of its uncomplicated network structure, strong self-learning ability and good generalization ability. The selection of RBF center point is always one of the main problems of RBF neural network algorithm. The existing selection methods of RBF center point are mainly based on K-means algorithm, but K-means clustering suffers from the low performance due to the low accurate initial value. In order to get the RBF neural network with better diagnosis effect, immune algorithm is used to calculate RBF center point, and then immune RBF neural network is obtained and used in power grid fault diagnosis. The numerical results verify the effectiveness and accuracy of the proposed algorithm for diagnosing the fault location of power grid.