震灾经济损失评估的遗传神经网络模型

地震灾害造成的直接经济损失与很多因素有关:致灾因子强度,主要包括地震震级、发震时间及地点、震源深度和地震动输入参数等;受灾体密度,主要包括衡量城市经济和社会发展水平的人口密度、城市密度、建筑物密度和财产密度等;城市抵抗地震灾害的能力.这里选取震级、地震动输入参数、人均国内生产总值GDP、受灾面积和灾区人口密度作为网络的输入节点,用直接经济损失率作为网络的输出节点,建立了基于遗传神经网络的震灾经济损失评估模型,对地震灾害所造成的经济损失进行评估,实例验证了该方法的有效性.

Genetic algorithm-neural network-based economic losses assessment of seismic disaster

Generally speaking,many factors affect economic losses of seismic disaster and they include:firstly,the disaster-causing factors such as the earthquake's magnitude, time, place and input parameters of ground motion; secondly, the density of disaster-affected body which is the indexes assessing the economy and development of cities such as population density, city density, building density and wealth density; lastly, the city's ability against earthquake disaster.This paper mainly selects the magnitude of earthquake, the input parameters of ground motion, per capita GDP, the disaster-suffered area and population density as input nodes while the rate of direct economic loss as output node.In this paper,the assessment model of seismic losses based on the genetic algorithm and artificial neural networks is made, which is used to assess the seismic losses.