基于主成分分析法的雷电灾害易损性区划综合评价模型

Comprehensive evaluation model of lightning vulnerability zoning based on principal component analysis

在研究区域雷电灾害易损性区划中,大多采用层次分析等方法来确定指标权重值,然而层次分析法在确定指标权重值时,人为主观因素较大,必然会给所建立的综合评估模型带来较大的误差.因此,利用主成分分析法,建立了雷电灾害易损性区划综合评价模型,从而很好地克服了层次分析等方法所带来的缺点,使得模型结果更接近于实际水平.利用海南省雷击大地密度、雷电灾害发生的频数、经济损失模数、生命易损模数以及人员伤亡频数等5个指标数据,采用所建立的雷电灾害易损性区划综合评价模型,对海南省雷电灾害易损性进行了实例分析.

The analytical hierarchy process （ AHP ） is frequently employed to determine the element weights in study of regional lightning vulnerability zoning. However,the scholar？s individual bias in judgment of option weight during AHP will introduce error to lightning risk assessment model.In order to avoid the potential error brought by AHP ,we use principal component analysis instead to establish a new type of lightning disaster vulnerability zoning model,which is then applied to analyze the lightning disaster vulnerability in Hainan province. Five indexes,inclu？ding thunderstrike density,lightning frequency,economic loss,life vulnerability module,and casualties frequency are chosen to zone Hainan province in lightning vulnerability.