水文时间序列相似性查询的分析与研究——以漯河站、何口站汛期降雨量相似性查询为例

Analysis and Study on Hydrological Time Series Similarity Search

水文时间序列相似性查询可用于雨洪过程预测、环境演变分析、水文过程规律分析等方面。最为直接的应用是回答防汛指挥中经常问到的"当前水文过程相当于历史上哪一时期的同类过程"等问题。由于水文数据数量大、类型复杂,仅使用水文学的理论和传统统计方法加以分析,往往因为计算能力、存储能力以及算法的复杂而造成预测时间较长。在水文领域引入数据仓库和数据挖掘的理论与技术,为解决水文科学研究面临的问题提供了新的思路。

Hydrological time series similarity search can be used for rainfall and flood forecasting,the analysis of environment evolvement and hydrological process,etc.The direct application is to answer such questions as ＂which hydrological process in history does the ongoing one correspond to？＂ The quantity of hydrological data is huge and its type is complexity.It is hard to analysis using only the methods of hydrological theories and traditional statistics,because of the limitation of computing ability,storage ability and algorithms.The introduction of data mining theories and technologies to hydrological area provides a new way to settle the problems that the hydrological science confronts with.