基于对象空间的地理场景表达模型与组织管理方法及应用

Organization and management of geographic scenes based on object space

信息智能化与时空大数据时代拓展了地理信息系统(GIS)的应用边界,给地理场景的描述与表达带来了机遇与挑战。现有的时空数据组织与空间认知模型主要以自下而上思维为主,存在对象抽象表达认知度低、实体对象分区组织割裂等局限性。特别是大数据背景下,地理场景表达中对象与知识的集成度较低、模型与数据的协同计算能力不足等问题更为突出。本文以莱布尼兹(Leibniz)相对时空观为理论基础,提出了对象空间的地理场景组织与管理方法。文章分析了地理场景的概念与发展历程,剖析了地理场景特点,提出了对象空间概念以及对象空间对地理场景的表达方式,建立了层次化对象组织模型、网络化的知识表达技术、按功能分类的模型组织方法,从而构建了“数据-知识-模型”高度集成与协同的地理场景建模理论与技术。结合山东高标准农田土壤含水量监测案例,验证了本文方法的实用性和可行性,为深化地理数据组织模型的理论研究和拓展示范应用场景提供了技术支撑。

The advent of information intelligence and spatiotemporal big data have significantly broadened the scope of application for geographic information systems(GIS),presenting both opportunities and challenges in the modeling of geographic scenarios.Current paradigms for organizing spatiotemporal data and conceptualizing spatial cognition predominantly rely on a bottom-up approach,which was demonstrated with limitation on low cognition and the fragmented representation of geospatial objects.This is a noteworthy research issue facing the Big Data era,namely the design of new representation models for the integration of objects and knowledge,as well as the collaborative computation of models and objects.This study,inspiring from Leibniz's relative spatiotemporal perspective,establishes an object space-based approach for organization and management of geographic scenarios.The concept of object space was proposed by reviewing the historical evolution of geographic scenarios representation model and literature work on mainstream research domain.Object space is the space of influence of an object,both the inner space of the object and the space of its surroundings.It includes pan spatiotemporal object,object space relationship,calculation and analysis process,which represents object,knowledge and model.For representing and management of object space,a hierarchical model was developed to organize pan spatiotemporal objects according to business requirement and spatial scale.Further,a network model was denoted to represent object space relationship and knowledge,in which node is the objects,and edge is the space relationship.Then,a model classification method based on functional and computational ability was used to organize calculation and analysis process models.Thus,a highly integrated and synergistic"data-knowledge-model"organization and management model was established.The proposed approach was applied in monitoring soil moisture in high-standard farmland in Shandong,which included 44 pan spatiotemporal objects,2 object space relationship network models and 5 calculation and analysis process models.The results demonstrated its efficacy and feasibility in designing and implementing high-standard farmland intelligent automatic irrigation and drainage systems,thereby offered technical support for advancing theoretical research and expanding practical application in geographic scenarios.