多粒度时空对象动态行为表达模型与方法研究

Research on Dynamic Behavior Expression Model and Method of Multi-Granularity Spatio-Temporal Objects

对具有动态变化特征的空间实体的描述,除了表达其内在属性外,还应包括动态行为。传统GIS把空间对象内在属性和动态行为分开建模本质上是一种基于地图的静态建模思想,而多粒度时空对象把行为能力作为对象的固有属性,是区别于其它对象的一个重要特征,使得全空间信息系统能够描述具有认知和行为能力的"活"的地理实体。本文针对具有规则化动态变化特征的时空实体的描述与表达,从多粒度时空对象的角度,阐述了多粒度时空对象动态行为的表达思路,提出了多粒度时空对象个体和对象世界的动态行为表达模型,为全面认识和表达空间对象动态变化提供了新的视角。针对动态行为多属性交错改变、多对象复杂联动、多过程动态自主的特点,为开展多粒度时空对象动态行为实际建模,本文归纳了动态行为改变内在属性的4种不同方式,并分别基于动力学、规则、数据挖掘和智能体4个方面对动态行为的表达方法进行了探讨。最后,结合实际应用进行了实证分析,结果表明对空间实体内在属性和动态行为一体建模,不再需要设计额外的数据模型,行为与对象不再分离;采用统一的动态行为表达模型,行为不仅与对象是解耦的,而且是通用的、可继承和复用的;对象不同行为的组合,可以描述空间实体复杂的动态变化的过程,为全空间信息系统中对象行为建模提供了有益经验和借鉴。

The description of spatial entities with dynamic change characteristics should include dynamic behavior as well as inherentl attributes. Traditional GIS models the inherent attributes and dynamic behaviors of spatial objects separately, which is essentially a static modeling idea based on map. However, the multigranularity spatio-temporal objects take the behavior ability as the inherent attributes of objects, which is an important feature different from other objects, which makes the Pan-Spatial Information System(PSIS) be able to describe the "living" geographical entities with cognitive and behavioral capabilities. This paper focuses on the description and expression of spatio-temporal entities with regularized dynamic change characteristics,expounds the expression ideas of dynamic behaviors of multi-granularity spatio-temporal objects from the perspective of multi-granularity spatio-temporal objects, and proposes a dynamic behavior expression model of multi-granularity spatio-temporal object individual and object world, which provides a new viewpoint for comprehensive understanding and expression of dynamic changes of spatial objects. According to the feature of multi-attribute staggered change, multi-object complex linkage, multi-process dynamic autonomy, In order to carry out the actual modeling of multi-granularity spatio-temporal object dynamic behavior, this paper summarizes four different ways of dynamic behavior changing inherent attributes, and presents the expression methods of dynamic behavior based on dynamics, rules, data mining and intelligence agent respectively It is discussed. Finally, an empirical analysis is carried out in combination with practical application. The results show that it is no longer necessary to design additional data models to integrated model the internal attributes and dynamic behaviors of spatial entities, and the behaviors and objects are no longer separated;using the unified dynamic behavior expression model, the behavior is not only decoupled from the object, but also universal, inheritable and reusable;the combination of different behaviors of objects can In order to describe the complex dynamic change process of spatial entities, it provides useful experience and reference for object behavior modeling in the PSIS.