论文部分内容阅读
大多数易损性评价方法,都需先定分级标准,且评价结果等级分辩率较粗.权重的确定受主观因素的影响,权重值因人而异,导致评价结果存在差异.投影寻踪方法(PP方法),可以克服以上的不足.基于实数编码的加速遗传算法可以解决PP方法计算量很大的问题.运用改进的遗传投影寻踪法,选择5个典型指标对湖南省洞庭湖区30个堤垸洪灾易损性进行了评价.结果表明:影响洞庭湖区洪涝易损性的主要因素是险堤长占总堤长的比例、控制点堤顶高程与历年最高水位平均差额.洞庭湖区洪涝易损性总体程度较高且无明显地区差异;重点堤垸与蓄洪堤垸易损性程度差异较大,前轻后重;易损性程度呈现出堤垸平均规模越小、易损性程度越高的规律.同时,与基于因素成对比较法的多因素综合评价结果进行对比,表明PP方法客观可靠,表征信息丰富,结果合理,具有有效性和通用性.
Most of the vulnerability assessment methods, we must first set the standard, and the evaluation results of the level of resolution is relatively thick.Determination of weight by the subjective factors, weight values vary, leading to differences in the evaluation results.Projection pursuit method (PP method) can overcome the above shortcomings.The accelerated genetic algorithm based on real number coding can solve the problem of large computational burden of PP method.Using the improved genetic projection pursuit method, The results show that the main factors that affect the vulnerability of flood in Dongting Lake area are the proportion of embankment in total length of embankment and the average difference between the height of embankment roof and the highest water level in past years. The overall degree of damage is relatively high and there is no obvious regional difference; the degree of vulnerability of embankments and flood embankments is different between before and after light; the degree of vulnerability shows the smaller the average embankment size and the greater the degree of vulnerability High rule.At the same time, compared with the result of multi-factor comprehensive evaluation based on the method of paired comparison, it shows that the PP method is objective and reliable, characterized by abundant information and reasonable result, which is effective and universal.