Research And Application On The Earthquake Disaster Loss Estimation Method Based On Computational Intelligence

Global earthquakes are mainly distributed in the circum-Pacific and HimalayanMediterranean seismic belts.China locates between these two seismic belts,and it is a country with frequent earthquakes.With the problems of dense population and poor earthquake resistance of buildings,earthquakes have become the natural disaster causing the death of most people in China.After the destructive earthquake occured,before the on-site investigation conducted,the predictions of the earthquake-affected area and the casualties of the earthquake victims are of great significance to the earthquake emergency response.Computational intelligence is a series of algorithms collectively known as being inspired by the intelligent behavior of the natural or biological world.Computational intelligence algorithms mainly include three parts: neural computation,evolutionary computation and fuzzy computation.These algorithms can solve many complex problems that cannot be solved by traditional methods and have been widely applied to scientific computation and engineering applications.In this thesis,two important methods of the BP neural network and the particle swarm optimization algorithm in the computational intelligence algorithm are applied to the two important applications of earthquake influencing scope prediction and earthquake personnel death prediction.The main work of the thesis are as follows:(1)This earthquake mortality model in Sichuan is studied in this thesis.The imbalance of regional economic development leads to large differences in the level of seismic fortification of buildings in different regions and between urban and rural areas,and there are often significant differences in the geological structure of different regions,and these differences can be reflected in historical seismic data.Thirty earthquake cases in Sichuan Province from 1970 to 2008 were collected.According to the USGS PAGER system empirical model of the intensity and earthquake mortality,the particle swarm optimization algorithm was used to find the optimal solution,and earthquake mortality at each intensity in the Sichuan region were obtained.Earthquake mortality model was validated by using the relevant data from the 2008 Wenchuan earthquake and the 2013 Lushan earthquake.(2)This seismic intensity attenuation fusion model is studied in this thesis.In order to further improve the prediction accuracy of the seismic intensity attenuation model,according to the characteristics of seismic tectonic environment and earthquake disaster distribution,Sichuan-Chongqing region is considered as a research area.250 seismic isoseismals from 110 earthquake cases was collected from Sichuan-Yunnan region.The predictive value of the long and short axis of the isoseismic line use seismic intensity attenuation model in western China and matrix intensity attenuation model as input.The final convergent model predicts the long and short axis of the isoseismic line as output.Using BP neural network to approximate the characteristics of nonlinear functions with arbitrary precision,the prediction results of the intensity attenuation model and the matrix attenuation model in the west of China are studied and fused,and the fusion prediction model suitable for the Sichuan-Yunnan region was obtained.(3)The seismic personnel mortality model and seismic intensity attenuation fusion model are designed and applied to the earthquake emergency assistant decision-making system.In this thesis,the intensity attenuation and fusion model of Sichuan and Yunnan area and the Sichuan earthquake mortality model are designed into two functional models and integrated into the earthquake emergency decision support system.The system can predict the lengths of the major and minor axes of seismic isolators at the first time after the earthquake,and thus obtain a preliminary scope of earthquake impact.The prediction of the number of deaths caused by the earthquake based on the obtained earthquake influence range,and the probability of the actual death of the earthquake in the interval,provide an auxiliary decision basis for the government emergency rescue.The system has been applied in Danling County,Sichuan Province.