The application of RBF network with ant colony algorithm in eddy current nondestructive detecting
Aihua Li, Xiaoyun Sun, Donghui Liu, Guoqing An · 2010
In eddy current nondestructive detecting, ant colony algorithm is adopted to select the centers of RBF network. The traditional clustering algorithm of RBF can't cluster much better, so the training speed and structure are limited. In this paper, ant colony algorithm with better optimal capability is used for optimize RBF network. The simulation result shows that ant colony algorithm has much optimal network than OLS algorithm. And the training speed and network robust are improved. It can be used for eddy current nondestructive detecting.