基于联系数集对势置信区间估计的旱灾风险分析模型

针对现有集对势方法难以进一步挖掘联系数中不确定性变化信息问题，提出将三元联系数减法集对势看作一个服从正态分布的随机变量，采用95%置信水平下置信区间的长度来反映不确定性大小，建立了基于联系数集对势置信区间估计的旱灾风险分析模型。以宿州市为例对构建的模型进行了验证，评价结果与现有的减法集对势、半偏减法集对势方法相一致，说明基于联系数集对势置信区间估计的旱灾风险分析模型合理有效；该模型得到的评价结果不是一个数值，而是一置信概率区间，提供了关于评价结果可靠性方面的信息，能够反映在多种不确定因素综合影响下的旱灾风险实际情况，在水资源保护系统不确定性分析评价中具有应用前景。

Drought risk analysis model based on confidence interval estimation of connection number set pair potential

To solve the problem that existing set pair potential methods hardly reflect the uncertainty of connection number, it was proposed to regard the subtraction set pair potential of ternary connection number as a random variable obeying a normal distribution, with the confidence interval at 95% confidence level used to reflect the uncertainty, and a drought risk analysis model based on confidence interval estimation of connection number set pair potential was established. Suzhou City was used as an example to verify the model, and the drought risk evaluation results of Suzhou City with the proposed model are consistent with those of subtraction set pair potential and semi-partial subtraction set pair potential methods, demonstrating that the drought risk analysis model based on the confidence interval estimation of connection number set pair potentials is reasonable and effective. The evaluation result obtained with this model is not a value, but a confidence probability interval, which provides more information on the reliability of the evaluation result, and can better reflect the actual situation of drought risk comprehensively affected by a variety of uncertainties. The proposed model in this study has application prospect in uncertainty analysis and evaluation problems of water resources protection systems.