Earthquake Prediction Based on Improved BP Neural Network

Artificial neural network is a new mathematical modeling, and the application field of which is very extensive. Uncertainty in the ground-motion amplitudes that will be realized from earthquakes plays a critical role in seismic hazardanalysis. The predisposing factors of earthquake are various and earthquake magnitude prediction is difficult. Considering the strong fault tolerant ability and fast velocity of prediction of neural networks, neural networks are investigated for predicting the magnitude of the largest seismic events and can get good prediction effect. An efficient method such as the LM-BP algorithm is introduced to predict the earthquake. The results show that the method by the LM-BP algorithm is effective and feasible for earthquake prediction. Compared with some other algorithms, the LM-BP neural algorithm has faster convergence and higher precision accuracy.