An Improved Direct Neural Network Approach To Flatness Pattern Recognition Baseed On GA-RBF

In this paper, an improved the radial basis function (RBF) neural network direct recognition approach to shape flatness pattern is proposed. The genetic algorithm (GA) is employed to obtain more optimal structure and initial parameters of RBF network. The new approach with the advantages of RBF, such as fast learning and high accuracy, is efficient and intelligent, it can not only effectively settle the problem of the different topologic configurations with changing strip widths but also improve practicability and precision. Compared to the improved direct recognition method with GA-BP, The simulation results show that the speed and accuracy of the flatness pattern recognition model based on GA-RBF are obviously improved.