基于可变模糊聚类迭代模型的溃坝生命损失预测

Forecasting Life Loss of Dam Break Based on Variable Fuzzy Clustering Iterative Model

针对我国溃坝相关资料缺乏、数据均匀连续性较差的现状,将可变模糊集理论引入到溃坝生命损失预测中,建立基于可变模糊聚类(VFC)循环迭代模型的人口死亡率预测模型。采用相邻模糊标度法确定指标权重,运用模型得到最优相对隶属度矩阵和最优聚类中心矩阵,结合欧氏距离,对样本进行二次筛选,再运用指数加权平均法,计算得到待估样本人口死亡率估计值。实例验算证明,该模型符合我国现状,具有一定的实用性和较高的精度。

In China,there is lack of relevant data of dam break and the data has poor uniform continuity at present.The variable fuzzy set theory is introduced to predict life loss of dam break.And the population mortality prediction model is established by using variable fuzzy clustering iterative method.Then adjacent fuzzy scale method is adopted to determine the index weight.At the same time optimal relative membership degree matrix and optimal clustering center matrix are obtained with the model.Based on the Euclid distance,it completes the secondary screening of samples.Finally,population mortality rate of samples is calculated by using exponential weighted average method.Example results show that the proposed model not only has a certain practical and high accuracy,but also fits for the current situation of China.