ANN Model for the Estimation of Life Casualties in Earthquake Engineering

With the development of regional economy and the progressing of citification, the population gets more and more centralized; this brings new issues to the research in earthquake-proofing. The research of prediction to life casualties related factors in an earthquake could give clear direction to the management of disaster-reduction work. In this paper, a BP neural network model is built up for the prediction to life casualties in an earthquake, which takes earthquake magnitude, depth of hypocenter, intensity of epicenter, level of preparedness, earthquake acceleration, population density, disaster forecasting as the key factors, employs 37 severe earthquake disasters happened in our country as samples for the training of the network. Examining result shows: this model fits for most cases of the earthquakes in our country.