| 摘要: |
| 【目的】 研究改进非参数解集模型在陕北地区月径流随机模拟中的适用性，为该区域水资源规划、开发利用和保护等提供决策依据。【方法】 采用非参数核密度估计模型模拟年径流，利用改进非参数解集模型将年径流分解为月径流，对月径流进行随机模拟，并探讨改进非参数解集模型中不同正定型矩阵的分解方法。【结果】 改进非参数解集模型考虑了分量与总量以及分量与各分量前期值的关系，克服了首尾自相关不一致的问题，对月径流实测序列的均值、均方差、Cv、Cs、最大值、最小值以及一、二阶自相关系数等统计特性保持较好，模拟效果比较理想。交口河站、张村驿站和绥德站S′矩阵为正定矩阵，可采用Cholesky分解法进行分解；刘家河站、志丹站、安塞站和杏河站S′矩阵为非正定矩阵，宜采用Schur分解法进行分解。【结论】 利用改进非参数解集模型进行月径流的随机模拟是可行的。 |
| 关键词: 月径流随机模拟 改进非参数解集模型 核密度估计模型 带宽系数 Cholesky分解 Schur分解 |
| DOI： |
| 分类号: |
| 基金项目:国家自然科学基金项目(50879070，50579065)；西北农林科技大学青年学术骨干支持计划和优秀博士论文基金项目(Z2007) |
|
| Application of improved nonparametric disaggregation model for stochastic simulation of monthly runoff |
|
|
| Abstract: |
| 【Objective】 Applicability of improved nonparametric disaggregation model is researched to simulate stochastic monthly runoff in Shaanbei area,which provides an important basis for regional water resources planning,development,utilization and protection.【Method】 This paper simulated annual runoff by kernel density estimation model,decomposed it into monthly runoff with improved nonparametric disaggregation model and did the stochastic simulation of monthly runoff,in addition,explored the decomposition method for the different positive type matrixes.【Result】 Improved nonparametric disaggregation model considers the relation of total and component,as well as the relation of component and its previous value,overcomes the discord of autocorrelation from beginning to end,and maintains good statistical properties of the measured sequences on mean,square deviation,Cv,Cs,maximum,minimum,first-order autocorrelation coefficient,et al.When S′ is positive definite matrix,Cholesky decomposition can be used at Jiaokouhe station,Zhangcunyi Station and Suide Station.While,when S′ is nonpositive definite matrix,Schur decomposition can be used at Liujiahe Station,Zhidan Station,Ansai Station and Xinghe Station.【Conclusion】 The result shows that improved nonparametric disaggregation model is practical to simulate stochastic monthly runoff. |
| Key words: simulation of monthly runoff improved nonparametric disaggregation model kernel density estimation model bandwidth coefficient cholesky decomposition schur decomposition |
