Application Of BP Neural Network In Rapid Assessment Of Earthquake Emergency Losses

In recent years, a lot of great earthquakes have occured in China, which caused serious losses and killed many persons.Therefore evaluating the earthquake death toll and economic losses is the necessary premise of earthquake emergency work,thus,the death toll and economic losses as data support of the earthquake emergency disposal measures,only to master the death toll and economic losses after the earthquakes,can we perform our earthquake emergency work efficiently.The main content of this paper is using the method of data mining to assess earthquake death toll and economic losses to better guide the earthquake emergency work.The method used in this paper is BP neural network model. The BP neural network model is an important model of artificial neural networks, which is widely used in many fields like pattern recognition and predictive assessment and so on.The death toll and economic losses caused by devastating earthquakes are decided by a series of different factors,Therefore,this article introduces the BP neural network model to assess the number of deaths and economic losses after earthquakes.The BP neural network model can store and learn a lot of input-output mapping modes without revealing the mathematical expression of this mapping,in addition,the BP artificial neural network transfer function between neurons are non-linear function, which makes this model has a comparative advantage in dealing with problems in nonlinear functions. For a three layer BP neural network model, which can approximate any non-linear function problems through many trains, thus, the BP neural network model can better depict the complex non-linear,multi-input-output and indefinite features in assessment of death toll and economic losses.The paper uses the BP neural network to establish the emergency earthquake fast loss assessment models,which calculate the number of deaths and economic losses after the earthquakes and establish the number of deaths prediction models and economic losses assessment model respectively.We can determine the relationship between the number of deaths and economic losses and natural factors by the model,s calculation,using the test data to validate the model in order to estimate the performance merits, the experimental results show that the BP neural network model is feasible to assess the death toll and economic losses,what,s more,the BPneural network model can effectively simulate the non-linear act in the death toll and economic losses.The result is more accurate.In addition,the result also shows many advantages of BP neural network model,including the objectivity of assessing results,the simplifying of the assessing model,the high speed of calculation and high precision and so on,so this method is valuable and has a bright future.