Research On Visualization Of Parallel Processing For Massive Floating Car Data

In the city public transportation taxi,the bus will produce the massive track data every day,faced with these massive data,because the traditional single node computing system's processing ability is limited,the massive data processing platform which is based on the parallel computation pattern emerges as a result,the mass data processing technology Hadoop supports the oversized data file,The data is processed and analyzed by the way of parallel processing.With the rapid development and application of urban vehicle location information acquisition technology,floating car position data has become an important data source of traffic state analysis.As an important part of city traffic floating car,taxi plays a very important role in city traffic.Taxi operators keep a large amount of Global Positioning System(GPS)data for taxis,and it is interesting to get useful information from this massive amount of GPS data.This paper uses Hadoop Big Data Platform and GIS Big Data Processing Module(GIS-Tools-for-Hadoop)launched by Esri Company to analyze the GPS data of taxi in Xi'anand excavate the hot spots of passengers,and use ArcGIS software to display them visually.Firstly,the thesis introduces Hadoop platform,the architecture and operation mechanism of HDFS(Hadoop Distributed File System),MapReduce(Hadoop Distributed Computing Framework),Hive(Hadoop Data Warehouse Tool),The composition and function of GIS-Tools-for-Hadoop module are introduced in detail.Then the method of building Hadoop massive data processing platform is introduced.Finally,GPS data analysis of taxi and visualization of passenger hotspot area are carried out.In the process of data analysis and visualization,map JSON files and taxi GPS data files are required and configured to use Hadoop geodata processing tool Geoprocessing-Tools-for-Hadoop to obtain JSON files for Xi'an city maps and to obtain taxi GPS data files from the Oracle database.Uploading the above two data files to HDFS,using Esri Geometry API to equalize the map area,dividing Xi'an city into multiple area blocks,matching the area blocks according to the latitude and longitude of the taxi GPS data,Through two MapReduce operations to obtain taxi passenger hot spot area and the number of taxi trips,finally use the Hadoop spatial information analysis framework Spatial-Framework-for-Hadoop in the Hive space query,the hotspot area display in ArcGIS software.