Research On Web-oriented Framework Construction For Visual Analytics Of Moving Objects

In the history known to mankind,the evolution of species and the development of human civilization have been greatly promoted by the improvement of mobility.In modern society,moving objects are widely found in all sectors of life.Important decisions in many fields,such as scientific research,transportation,commerce and so on,depend on the understanding of moving objects.Advances in technologies such as global positioning,wireless transmission,and data storage have laid the technical foundation for the acquisition and storage of moving objects.Therefore,how to use these data effectively is a difficult problem that needs to be solved urgently in the current society.Currently,data mining technology explores the information hidden in data through various algorithms,but it has the disadvantage of high professionalism and cannot be extended effectively.Traditional data analysis methods include various data visualization and statistical analysis platforms,often lacking proprietary analysis components and having a relatively single function.Visual analytics combines data visualization and data analysis algorithms to guide users to explore data through an interactive analysis process,which can fully exert human cognitive advantages and computer processing capabilities.However,most of the current visual analysis tools focus on the analysis of geospatial data or domainspecific data,and lack a universal effective analysis platform for moving objects.And most of platforms are native desktop applications which are difficult to install and maintain,most of them are complex to operate and cannot meet the needs of popular visual analysis.According to multi-dimensional,space-time and multi-perspective and other characteristics of moving objects,a Web-oriented visual analytics framework for moving objects is designed in this thesis on the foundation of fully researching the characteristics of moving objects and the existing visual analytics framework.The framework consists of three main parts: data management module,data visualization module and visual analysis module,which has the characteristics of universality and modularization.The main work of this thesis is as follows:(1)The relevant theories of visual analysis of mobile objects are fully studied in this thesis,then the characteristics and problems of existing platforms are summarized.The characteristics of mobile objects are analyzed from the perspective of visual analysis,and three visual perspectives of mobile object analysis are put forward,then the conceptual model of visual analysis core computing process and data transformation pipeline for mobile objects are refined.In the last,the overall framework of visual analysis framework for moving object is designed,which provides a theoretical basis and conceptual framework for the implementation of key technologies in the future.(2)The data management model of the visual analysis framework is designed on the foundation of researching the data management model in Open Layers.This data management model is designed with modular multi-source data import to suit different needs,such as a data management part of multi-view collaborative change for a variety of visual analysis tasks and a data service and data storage part for complex data processing and data sharing requirements.(3)The visualization part of style and logic separation is designed based on the study of the existing framework visualization model.This visualization model realizes data-driven visual presentation,solves the single and messy visualization process of Open Layers,and designs an interactive visualization module according to the characteristics of moving objects.(4)The common methods of multi-view visual analytics of moving objects are studied extensively,and a visual analysis module that can quickly add and delete components is implemented.The visual analysis components is divided into two types in this thesis: computational and interactive,and realize some representative analysis components.Finally,the prototype system of the moving objects visual analytics framework was built based on the above design,and accessed various types of data to test the function of the system.The test results show that the visual analytics framework designed in this thesis is rich in functions and easy to use.The modular and low-coupling design between different parts enables different types of data to be quickly imported and visual analysis customized.Various visual methods and visual analysis components can provide flexible and effective multi-perspective visual analysis for moving objects.