Research And Design Of Large Vehicle Sensing Data Distributed Manegement Platform

In urban computing system,it is necessary to analyze,mine and fuse large heterogeneous data to solve various problems in the city and provide related intelligent applications and services.Vehicle sensing data is important in urban computing system,which has the characteristics of mass,high speed and diversity.Management for vehicle sensing data faces many challenges.This paper purposes a management platform for vehicle sensing data,which provides efficient data preprocessing and mass distributed data management functions.Aiming at the data characteristics of vehicle sensing data and the demand of urban computing applications and services,a series of algorithm design and engineering optimization design are proposed.Firstly,this paper analyses the research background and main challenges of the massive vehicle sensing data distributed management platform,and clarifies the scenarios of the platform in urban computing system.Then,a complete vehicle sensing data specification is proposed based on the characteristics of vehicle sensing data and platform application requirements.Further,according to the data specification and application requirements,a series of data preprocessing rules and algorithms at different levels are proposed.According to the idea of information theory,a scalable trajectory compression algorithm HESAVE and a trajectory map matching algorithm InfoMatch are proposed,which have certain improvement compared with the relative algorithms.After that,the requirement analysis,design and feasibility validation of the platform are carried out,and a reasonable software framework are designed,as well as software detial design of key function modules such as data preprocessing,data management and unified query interface.Aiming at the reliability,security,real-time and scalability of vehicle sensor data management platform,the paper proposes some optimization plans.Finally,the feasibility of the platform is verified.The results show that the platform can meet the requirements of vehicle sensor data management in urban computing system.