针对溃坝生命损失历史资料缺乏且影响因素复杂的特点,鉴于支持向量机算法的诸多优势,建立了基于支持向量机的溃坝生命损失评估模型。将溃坝生命损失定义为风险人口与生命损失率的乘积,分析了生命损失的评估指标,建立了综合评估体系。根据我国已溃决大坝的历史资料,构建了学习样本。留一法交互校验结果表明,该评估方法具有一定的精度。
Considering shortage of historical data and many complicated influence factors about loss of life caused by dam breach, an evaluation model of loss-of-life caused by dam breach based on support vector machine is put forward. In this paper, loss of life caused by dam breach is defined as the product of people at risk and death rate. Five important factors affecting loss of life are analyzed. The evaluation index system is established. Learning samples are constructed on the basis of historical data of eight breached dams in China. Leave one out cross validation shows accuracy of the evaluation method.