Seismic Pattern Recognition of Nuclear Explosion Based on Generalization Learning Algorithm of BP Network and Genetic Algorithm

During the pattern recognition using BP neural network, the generalization performance often becomes poor. To improve the generalization performance of BP Network, a novel BP network generalization learning algorithm based on suboptimal criterion of fitting error of random assistant samples is presented. And we apply this algorithm to the classification of underground nuclear explosion earthquake events and natural earthquake events. Experimental results indicate that this method is effective and can improve the identification rate of underground nuclear explosions and natural earthquakes.