基于模糊模式识别的径流特性分析

为了探讨河流天然径流的演变规律，引入模糊模式识别的理论和方法对流量序列进行研究，给出了基于模糊聚类的径流丰枯特性分析的方法；在对径流丰、中、枯聚类基础上提出了基于模糊模式识别的多维丰枯周期特性分析方法。以嘉陵江支流白龙江下游的三磊坝水文站多年月平均流量数据为例，得出了该站径流各年的丰枯特点、丰枯周期及相应的隶属度。该方法的多维径流特性分析结果与分析尺度和数据维数有关。

Runoff Characteristic Analysis by Fuzzy Pattern Recognition

In order to study evolution of natural runoff, this paper introduces fuzzy pattern recognition theory and method to research runoff series, presents the method of analyzing runoff characteristic with clustering of high, middle and low levels of runoff.This paper also gives the method of analyzing multidimensional high and low states periodicity of natural runoff based on fuzzy recognition arithmetic.The runoff characteristic, its periodicity and subjective of high and low levels of annul runoff from monthly average runoff of Sanleiba hydrology station in lower Bailongjiang River, a branch of Jialingjiang River are giver as example. The results of multidimensional runoff characteristic analysis are related to analysis scale and data dimension.