基于熵权的模糊聚类模型在山区融雪洪水产流类型中的分类应用

发布时间：2018-05-31 23:40

本文选题：军塘湖 + 融雪洪水　；参考：《中国农村水利水电》2017年05期

【摘要】：山区融雪洪水流量过程因洪水成因不同而异,但又表现出自身的规律性。将洪水产流类型分为若干类能够更好的掌握洪水机理和进行洪水预报。运用熵权确定影响因素的权重,并利用模糊C均值聚类模式识别分类方法得出洪水样本的自然分类。实验区位于新疆天山北坡山区地带的军塘湖流域。选取了12场洪水过程的观测资料,借助试验区架设的自动气象观测站的气象资料,积雪样方观测资料,将洪水的产流类型分为:超渗产流型,超渗蓄产流型,蓄满产流型。研究表明:(1)熵权能够有效地对影响洪水的各要素赋权,从而让模糊C均值聚类分析能够区分每个洪水样本的类型。(2)该方法能够准确、迅速的判断洪水产流类型,能有效地提高洪水预报精度。开展融雪洪水过程线分类研究,对于融雪洪水预报与预警有重要的理论和实际意义。
[Abstract]:The flow process of snowmelt flood in mountainous area varies with the cause of flood, but it also shows its own regularity. The flood mechanism and flood forecast can be better understood by classifying flood runoff types into several types. The weight of influencing factors is determined by entropy weight, and the natural classification of flood samples is obtained by using fuzzy C-means clustering pattern recognition and classification method. The experimental area is located in the Juntang Lake basin on the northern slope of Tianshan Mountain, Xinjiang. Based on the observation data of 12 flood processes and the meteorological data of the automatic meteorological observation station set up in the experimental area and the snow sample observation data, the runoff generation types of the flood are divided into three types: the superosmotic runoff generation type, the superosmotic storage current generation type and the full production flow type. The research shows that the entropy weight can effectively weight the factors that affect the flood, so that the fuzzy C-means clustering analysis can distinguish the type of each flood sample. 2) this method can accurately and quickly judge the flood runoff type. It can effectively improve the accuracy of flood forecast. It is of great theoretical and practical significance to study the classification of snowmelt flood process line for forecast and early warning of snowmelt flood.
【作者单位】：新疆大学资源与环境科学学院;新疆大学干旱生态环境研究所;新疆大学绿洲生态教育部重点实验室;
【基金】：国家自然科学基金面上项目“天山北坡典型流域积雪-冻土水热耦合中融水产流机制研究”(41171023)
【分类号】：P333.2;TV122

【相似文献】