Research And Optimization Of Data Warehouse In Vehicle Monitoring System

With the fast progress of technology and the rapid rise of economy, China’s automotive industry has been developing at a high speed. The increasing number of vehicles brings a series of problems, such as car crime, traffic congestion, and frequent traffic accidents, which not only bring great inconvenience to people’s daily life, but also affect the development of the society. To solve these problems, the idea that combines vehicles with Internet technology is proposed, which is called Intelligent Transportation.Vehicle Monitoring System, as the main topic of Intelligent Transportation research, is the effective means to solve some of the problems above. This paper designs a monitoring center, which is easy to split and combine, and adapt to a variety of applications. The design principles, system structure, technical implementation and the main functions are introduced in detail. As the system scales and time goes by, the monitoring center will accumulate a large number of valuable vehicle-location data and alarm data, which provide rich resources for users to analyze driving records and alarm information. To make full use of these historical data, Data Warehouse technology is introduced to Vehicle Monitoring System, and Hive Framework is used to optimize the implementation.First, this paper describes the design principles and structure model of Data Warehouse, and introduces the definition, classification and function of Meta Data. Next, on the basis of detailed analysis of original data, this paper designs the data model of Data Warehouse, abstracts four themes, including drivers, vehicles, driving records and alarm records, and designs the corresponding fact tables, dimension tables, the fact-dimension relation model and physical model according to each theme. Then this paper chooses the level of granularity according to the specific needs of business, and designs ETL process based on Hive framework. Finally, on the basis of Hadoop and Hive, this paper builds an experimental platform of Data Warehouse, which will provide statistical analysis and decision support for Executive Officers. The effect of this platform in handling big data is also examined through the test.