基于熵权法的模糊物元模型在山洪防御能力评估中的应用

Application of entropy weight method-based fuzzy matter-element model to assessment on mountain torrent prevention capability

针对我国山洪灾害评估工作的需求,运用信息熵与模糊物元理论开展区域山洪防御能力评估研究。遴选山洪防御能力中关于风险性、基础应灾能力、监测预警能力和群测群防能力的16项指标构建评估体系,引用熵权法计算指标权重,采用模糊物元方法建立山洪防御能力评估模型。以历史灾情为基础验证资料,分析风险层指标的模型输出结果,后推广至山洪防御能力综合评估中,并与层次分析法评估结果进行对比。结果表明:德宏州各县域的山洪防御能力由高到低的排序为瑞丽市、陇川县、梁河县、盈江县、芒市,其中仅芒市山洪防御能力等级为低等水平,其余县域为中等水平,并与层次分析法评估结果一致。该模型计算简单、结果合理,实用性较强,适用于多因子评估问题。研究成果为我国山洪防御能力评估提供了新方法。

Aiming at the demand of the assessment made on mountain torrent disaster in China,the study on the assessment of reginal mountain torrent prevention capability is made with the theories of information entropy and fuzzy matter-element herein;for which 16 indexes for the mountain torrent prevention capability related to risk,basic disaster response capability,monitoring and early warning capability and mass observation and mass preparedness capability are selected for establishing an assessment system,and then a model for evaluating mountain torrent disaster is built up with fuzzy matter-element method through the calculation of the index weight made by entropy weight method.By taking the relevant historical disasters as the basic verification data,the model output results of risk layer indexes are analyzed and then generalized to the comprehensive assessment of mountain torrent prevention,which are compared with the evaluating results from analytic hierarchy process(AHP)as well.The result shows that the sequence from high to low of the mountain torrent prevention capabilities of all the counties within Dehong Prefecture are Ruili City,Longchuan County,Lianghe County,Yingjiang County and Mangshi;in which only the mountain torrent prevention capability of Msngshi is on the low level and the mountain torrent prevention capabilities of all the other counties are on the mid-level;which are coincided with those from analytic hierarchy process(AHP).The calculation of this model is simple with reasonable result and stronger practicability,thus is applicable to the relevant multi-factor assessment.The study result provides a new method for the assessment of mountain torrent prevention capability in China.