Research On Data Analysis And Visualization Technology Of Electric Bicycle Track

With the development of low-carbon environmental protection society and the continuous expansion of urban scale,electric bicycles have become one of the important modes of travel in many cities in China due to their low cost,flexible operation and convenient riding.In recent years,with the rapid development of the electric bicycle industry,the related traffic and social security issues have emerged in an endless stream.In order to solve the contradiction between the increasing number of electric bicycles and limited roads,government departments need to fully understand the behavior characteristics of electric bicycle groups in order to strengthen the management of electric bicycles,thereby improving the service level and management efficiency of urban systems.The thesis carries out systematic research work based on massive electric bicycle trajectory data.In-depth analysis of the behavior of electric bike users and the construction of a feature-rich visual interaction system.The specific research content is as follows:The thesis is based on the big data platform to clean the electric bicycle data.For the trajectory data after cleaning,a speed-based time clustering algorithm is designed to extract the user’s stay point during the movement.The user is based on the Spark-based K-means++algorithm.The stay point is clustered to obtain the user hotspot stay area.In order to obtain the user’s place of residence information,different clustering algorithms are designed for users with different professional backgrounds to automatically and effectively mine the user’s place of residence information.Based on the data analysis,the thesis also builds a Web-based visualization system based on the big data platform,which is convenient for people to intuitively analyze and understand the electric bicycle data.The system includes the electric bicycle data map and the trajectory movement module.The front-end page often needs to retrieve the electric bicycle data stored in HBase according to the specified field or several fields.In order to improve the speed of the query,the thesis proposes a method based on Phoenix to implement HBase secondary index.finally,test the performance of Hfiase and Phoenix integration.Experiments show that the data query speed will be significantly improved after the secondary index is built on the HBase table data.The thesis combines big data with WEB visualization system,analyses the behavior of users of electric bicycle.According to the data characteristics analyzed,a visualization system of electric bicycle based on large data platform is constructed.Phoenix is used to create a secondary index for HBase,which improves the query speed of the visualization system.It provides a possibility for the practical application of this research work.