Neural network with adaptive genetic algorithm for eddy current nondestructive testing
Sun Xiaoyun, Liu Donghui, Kai Zhang, Liwei Guo, Zhen Ran, Jianye Liu · 2004
For eddy current nondestructive testing (ECNDT), adaptive genetic algorithm (GA) is adopted, which can overcome the disadvantages of back propagation (BP) artificial neural network (ANN), such as a possibility of being trapped on locally minimum value. Moreover, GA operators are selected by adaptive algorithm to overcome the prematurity. Compared with BP-ANN, the convergence precision and generalization of GA-ANN are improved remarkably.