面向水系河网的建模与索引方法研究

Research on the Modeling and Indexing Method for River Network

针对水系河网的非欧时空特征限制了数据查询效率,提出了一种面向水系河网的建模与索引方法。该方法首先建立水系河网的多尺度图论化模型,在此基础上定义河段的概念,并建立基于河段的流域水系河网索引(BRR-Tree)实现水利领域特征的表达,用以实现水利领域语义查询请求的表达;然后设计基于BRR-Tree的水系河网查询算法,并按照水系河网进行数据访问需求转换为欧氏空间访问需求。实验表明该方法能够有效地支持水系河网分析,提高水利领域数据的访问效率。

The water conservancy data appears as the sets of river-network-based time series, the data queries usually need to obtain the corresponding data according to the spatial distribution of the river network＇s structure. Regarding the problem that the river network＇s non-Euclidean spatial features limit the data query efficiency, this paper builds the river network＇s multi-scale graph theory model based on the quotient space theory first, then selects the river segments as the query units to construct the river network＇s indexes basin river R-tree（BRR-Tree）, designs the typical query algorithm of the river network. The algorithm converts the non-Euclidean data access requests into the ones meeting the Euclidean spatial features. The experiments show that the method proposed in this paper can effectively support the river network＇s analysis and improve data access efficiency in water conservancy field.