The application of RBF networks based on artificial immune algorithm in the performance prediction of steel bars

This work presents a novel radial basis function (RBF) neural network model based on immune recognition principle. This model can choose the number and location of the hidden layer centers by applying the principles of recognition, memory, learning and self-organized adjustment, and can determine the weights of the output layer by adopting least square algorithm. This novel model is applied to predict the performances of hot-rolled steel bars, and it achieves good effect. Simulation results show that this model proposed in the paper has the advantages of less computation and higher precision, compared with the k-means algorithm.