| 摘要: |
| 【目的】建立合理的多属性水文预报组合决策方法，为提高水文预报的精度提供参考。【方法】根据水文预报的特点，提取了水文预报中的4个决策属性指标，利用主观权重和熵权结合的方法确定属性指标的综合权重。将模糊物元理论和信息熵引入水文预报研究领域，利用模糊物元理论处理水文预报中的模糊信息，然后结合信息熵对属性进行集结，按照集结后方案的相对贴近度进行择优排序，建立了基于模糊物元的多属性水文预报组合决策方法。【结果】应用该组合决策方法对断面各时间点的预报模型进行择优，得到了断面在某一时间段与实测值更为接近的预报值。实例计算结果表明，该组合预报方法能够择优得到各方案预报值相对误差较小的最优值，有效地提高了水文预报精度，将预报结果至少提升了1个等级。【结论】所构建的基于模糊物元的多属性水文预报组合决策方法，具有较强的合理性及较广的适用性，为水文预报精度的提高提供了一条有效途径。 |
| 关键词: 水文预报 预报模型 组合决策 模糊物元 |
| DOI： |
| 分类号: |
| 基金项目:国家高技术研究发展计划（863）项目“面向水利信息化的高性能计算与网格应用”（2006AA01A126）；国家自然科学基金项目“面向防汛抗旱会商的综合集成平台及知识服务模式研究”（50979088） |
|
| Hydrological forecasting decision combination based on multi-attribute fuzzy matter element |
|
|
| Abstract: |
| 【Objective】The study was to establish a reasonable decision combination for multiple attribute hydrological forecasting,so as to provide important decision basis to improve forecast accuracy.【Method】The study described the extraction of its attributes of hydrological forecasting.Synthesis weight was determined by subjective weight of experts and entropy weight.The fuzzy matter-element and comentropy were introduced into hydrological forecasting research field,and the fuzzy matter-element theory was used to deal with fuzzy information of hydrological forecasting,and the effective aggregation of attribute made use of comentropy.The scheme’s preferred order according to the relative closeness,hydrological forecasting decision combination based on multi-attribute fuzzy matter-element was established.【Result】This study used the combined method to pick the best hydrological forecasting model of the same section in different time points.We obtained the forecasting with measured time period much closer to the prediction value.The results showed that the combine method can forecast optimal value of the small relative error for merit-based scheme.It could improve the accuracy of forecasting results effectively,and improve the level of forecasting value.【Conclusion】The hydrological forecasting decision combination based on multi-attribute fuzzy matter element had strong rationality and applicability.An effective way was provided to improve forecast accuracy of hydrological forecasting. |
| Key words: hydrological forecasting forecast model decision combination fuzzy matter element |
