Assessing Seismic Fire in City with Neural Networks

摘要

Cities are confronted with the threat of the earthquakes disastrous seriously. Among the secondary disastrous after earthquakes, fire is usually the most dangerous. It's import to research and predict the fire after earthquakes. Based upon the theory of B-P arithmetic, the fire loss after earthquakes was provided using neural networks approach. A number of data that include sequence information and make certain the model of neural networks by calculating step by step so as to achieve the prediction of the fire losses after earthquakes, taking a practical project for example, giving an analysis example. The result can be used to evaluate the post-earthquake fire hazard of urban residential area and prepare for hazard and make a decision for fire protection aid based upon these.