Design And Implementation Of Big Data Acquisition And Analysis System For Intelligent Traffic Flow

With the rapid growth of urban road traffic flow and the lack of certain traffic planning and management experience,traffic congestion,traffic accidents and environmental pollution have become more and more serious.In this case,being able to understand traffic conditions in real time and plan traffic routes reasonably is particularly critical to solving the above problems.At present,although navigation software such as Baidu map and Gaode map can display real-time road conditions,their own role positioning problem makes them only provide a relatively single function.In addition,the massive traffic data collection to the back-end system caused the economic loss of storage equipment and the technical constraints of data analysis and processing,which also needs new technical ideas and methods to solve.In summary,this paper designs and develops an intelligent transportation vehicle flow big data collection and analysis system,studies the link transmission success probability theory based on the cellular network,and applies it to the vehicle flow data collection module,and then uses big data technology to complete data processing and realize the visual operation.The specific research work is as follows:1.Interfering signals are the main factors limiting network performance.An easy-to-handle SINR analysis framework based on cellular network is proposed.According to the theory of random geometric mathematics,the accurate expression of link transmission success probability under interference-free,exponential fading and Rayleigh fading is derived.On this basis,through simulation,the effects of multiple system parameters on the success probability are analyzed,and the results of the success probability and the area spectral efficiency in several interference environments are compared.Compared with the traditional cellular network SINR analysis method,this method does not require meshing,which reduces the computational complexity and improves the performance of the cellular network.2.Based on Apache Hadoop,Apache Spark and related ecosystem component technologies,a core engine for distributed traffic big data collection and analysis is designed,the overall technical architecture of the engine is designed,and data collection,cleaning,and storage are performed And offline and real-time computing and other technical issues Among them,the success probability provides a theoretical basis for data screening in the system data collection module,and solves the problems of data redundancy and consistency Use HDFS and HBase technology to solve the problem of distributed storage of massive data.The data cleaning solution is implemented through the MapReduce computing framework?The offline and real-time big data processing functional components are designed through Spark SQL and Spark Streaming3.Introduced the detailed implementation process of the system,the first is the system requirements analysis,functional architecture design,database design.Then,using Spring boot,Layui and ECharts technology to develop the web-side visual system operation interface,and display the system's functional modules in a graphical manner.Finally,using a third-party software to build a complete cluster performance monitoring system,and through comparative experiments to prove the feasibility of the cluster and good computing performance.