Relative Space-based Spatio-temporal Modeling And Analysis For The Group Dynamics Of Moving Objects

The relative motion of moving objects is an essential research topic in geographical information science(GIScience),which supports the innovation of geodatabases,spatial indexing,and geospatial services.This analysis is very popular in the domains of urban governance,transportation engineering,logistics and geospatial information services for individuals or industrials.Importantly,data models of moving objects are one of the most crucial approaches to support the analysis for dynamic relative motion between moving objects,even in the age of big data and cloud computing.Traditional geographic information systems(GIS)usually organize moving objects as point objects in absolute coordinated space.In absolute space,geocoded locations are bound to previously existing geometry and topology relationships among the corresponding points in the space.The derivation of relative motions among moving objects is not efficient because of the additional geo-computation of transformation between absolute space and relative space.This is also largely the reason why the proper integration of GIS with geographical models,especially those describing social phenomena,continues to be so difficult.Moreover,the prediction method,which is based on these absolute based-models,cannot interact with other moving object directly for the absolute space represent the moving object as an independent entity.In short,the current models are lack of understanding object's behavior and dynamic relative relationship with other surrounding objects,because few models support dynamic relative relationship operation powerful.Therefore,these absolute based-models require a large amount of computation to support the dynamic relationship analysis from the perspectives of individual or groups of moving objects.Actually,relative space is an instinctual approach to moving objects and is a powerful theoretic framework to represent the surrounding dynamics and motion trends of moving objects in nearby crowds.In relative space,relative dynamic relationships between moving objects are easy to build independently of whether they can be geocoded by coordinates.Importantly,the analysis of moving objects in relative space could easily follow instinctual requirements.Therefore,the motivation of this paper is to create a relative space-based GIS data model of moving objects and propose some basic GIS operators for analyzing moving objects,which changes the analysis of current absolute space-based GIS models and facilitates the efficient computation of real-time relative relationship dynamics,such as the surrounding dynamics and motion trends of crowds near moving objects.Based on the elative space-based GIS data model of moving objects,a framework is proposed for describe and analysis the objects' movement in relative space.In addition,a predictive model of movement pattern dynamics of moving objects in relative space is proposed by considering object's behavior tendency and surrounding dynamic scenario on the basis on the describtion and analysis framework in relative space.More specifically,the contents of this paper are:(1)Proposing a relative space-based GIS data model of moving objects.This paper presents a novel model for storing,managing and analyzing moving objects based on their relationships in relative space by extending the space-time cube model,which aims to increase the efficiency of dynamic relative relationship analysis.In this model,the X and Y in space-time cube model are substituted as the reference and target objects,and the space-time bin in space-time cube are extended to store the data structure related to relative relationship(relative distance and relative angle)among moving objects.Based on the structure of data model,this paper introduces six classes of operators to support the orgnazition,storage and analysis of moving objects.(2)Introducing a framework for relative dynamic description and analysis.In absolute space,the movement of moving objects are recorded by coornates,and their moving states are described by the change of coornates in time series,such as speed and acceleration.This paper proposes some new quantities to describe the moving states in relative space according to the change of the relative relationships.The relative dynamic is defined basned on those new quantities for describing moving features in relative space.Those quantities twelve relative dynamic patterns between two objects are to characterize moving object's behavior in relative space based on their moving features in relative space.Relative dynamic scenario is defined to model object's surrounding dynamic scenario in relative space.Moreover,a description framework is built,which is consisting of relative dynamic,relative dynamic pattern and relative dynamic scenario.Based on description framework,four algorithms are proposed to retrieve relative relationships and analysis the individual and group behavior in relative space.The results of experiments show the proposed method saves about 50%-95% execution time compred with the method in absolute space.(3)Presenting an improved Relative Pattern-Movement Dynamic Bayesian Network(RMP-DBN).Owing to weak support of pattern and dynamics of the relative movements,the current prediction models always reprensent the moving objects as independent maneuvering entities,which don't consider their behavior tendency and surrounding dynamic scenario.Disregarding these factors may lead to erroneous interpretations of the situations,and decrease the prediction precious.This paper builds an improved Relative Pattern-Movement Dynamic Bayesian Network(RMP-DBN)to compute the possibility of relative movement and then to predict the future relative movement,which considers the constraints from object's behavior tendency and surrounding relative dynamic scenario.The improved RMP-DBN is organized as two-layer model.The first layer computes the probability distributions of different future relative movement patterns and the second layer predicts future relative movement based on predicted relative movement pattern from the first layer.The accuracy of RMP-DBN is more than 95%.