摘要
地理网络建模强调的是对逻辑网络中的节点和联线的网络拓扑和语义操作过程。而传统 GIS 矢量数据表达所采用的平面图原子级弧段-节点数据模型强调的是几何信息的表达。对此,作者讨论了如何将基于特征的 GIS 概念建模方法引入地理网络表达中,克服平面图数据模型的不足之处,发展基于特征的地理网络数据模型,并着重阐述了逻辑网络模型的特点和应用特征。在地理网络的表达方法中,作者分析了面向对象方法和数据库技术相结合的途径,比较了传统关系模型和面向对象模型在地理网络表达中的差异,以及面向对象方法在地理网络路径系统表达和动态分段及其网络连通性规则表达中的优势所在。
Geographical network modeling emphasizes the network topology between the nodes and links withinthe logical networks and the semantic manipulation process.But the classical GIS vector data representation isbased on planar graph primitive arc-node data models,which emphasizes geometrical representation.In this pa-per,the authors introduce a feature based GIS conceptual modeling method to break through the restriction ofplanar enforcement.To this end,a feature based geographical network model is developed and the characteris-tics and application issues are discussed.The author separates the network topology from the geometrical dataand argues that network topology is not always dependant on geometrical data although different copies of net-work topologies can be derived from a same geometrical network.Geographical network topology is differentfrom geometrical topology.Most of network modeling and analysis is based on derived network topology.For theimplementation of geographical network,the integration of object oriented methods and database technologies isfirstly discussed,the difference between classical relational models and object oriented models is then comparedin geographical network representation,and the advantages of object oriented methods for geographical networkroute system representation,dynamic segmentation,and network connection rules are analyzed lastly.It is ar-gued that the feature based object oriented approaches are powerful for complex network modeling and deservesdetailed research and development in network operators,extended structural querying language,and multi-scaledatabases implementation.
出处
《地球信息科学》
CSCD
2004年第3期72-78,共7页
Geo-information Science
基金
国家自然科学基金项目(40201043)
关键词
基于特征
面向对象
地理网络
数据模型
feature based
object oriented
geographical networks
data modeling