| 摘要: |
| 【目的】研究基于DS证据理论的多模型洪水预报方案优选模式，以丰富多模型洪水预报方法。【方法】建立了多模型洪水预报情形下的优选指标体系，运用基于证据可信度的DS证据理论合成方法进行洪水预报方案的优选，采用新安江模型、萨克拉门托模型、水箱模型和陕北模型进行多模型洪水预报的优选，并以东洋河为例进行模拟验证。【结果】在对东洋河洪水进行预报时,新安江模型、萨克拉门托模型、水箱模型和陕北模型的可信度（m(Mi)）分别为0.758 7，0.906 6，0.838 4和0.859 1，多模型洪水预报方案由优到劣的次序依次为：萨克拉门托模型、陕北模型、水箱模型、新安江模型，因此认为采用萨克拉门托模型对应的预报方案为最优预报方案。【结论】基于DS证据理论的多模型洪水预报方案优选可以为决策者提供更为科学的决策，进一步丰富和完善了多模型洪水预报方法。 |
| 关键词: 洪水预报 多模型 方案优选 DS证据理论 指标体系 |
| DOI： |
| 分类号: |
| 基金项目:国家自然科学基金项目（51109175）；水利部公益性行业科研专项（201001011）；中国博士后科学基金项目（20110490170）；陕西省科技攻关计划项目（2012K06-20）；陕西省教育厅省级重点实验室项目（12JS066）；陕西省教育厅科研计划项目（12JK0472） |
|
| DS evidence theory based scheme optimization of multi-model flood forecasting |
|
|
| Abstract: |
| 【Objective】This study optimized the multi-model flood forecasting base on the DS evidence theory.【Method】Multi-model optimization index system was established,and DS evidence theory was used to conduct flood forecasting scheme optimization.Xin’anjiang Model,Sacramento Model,Tank Model and Shaanbei Model were used as examples and a Dongyang River case study was performed.【Result】The credibility values (m(Mi)) of Xin’anjiang Model,Sacramento Model,Tank Model and Shaanbei Model respectively were:0.758 7,0.906 6,0.838 4 and 0.859 1,respectively.Sacramento model was the selected best forecasting scheme.【Conclusion】DS evidence theory based scheme optimization of multi-model flood forecasting offered policymakers more scientific decision-making information and enriched and perfected the multi-model flood forecasting method. |
| Key words: flood forecasting multi-model scheme optimization DS evidence theory index system |
