Skip to Main Content
In this paper, RBF neural network (RBFNN) based on genetic algorithm (GA) optimization is presented to deal with line losses calculation for power distribution networks. The centers and widths of hidden layer and the weights of the output layer are coded into one chromosome. It strengthens the cooperation between the hidden layer and the output layer, and avoids the risk of getting stuck into a local minimum. The simulation results indicate that the method presented has the advantages of a simple model, speedy learning and high precision.