Research On Intelligent Analysis Technology Of Multi-source And Heterogeneous Traffic Big Data

With the continuous development of detection technology and information sharing technology,the data acquisition equipment for intelligent transportation system stores massive urban traffic data.How to analyze multi-source and heterogeneous traffic data and study the factors affecting its spatiotemporal distribution characteristics has become the focus of current research.However,traditional data analysis methods can't meet the urgent demand of big data.Therefore,intelligent analysis technology is needed to effectively mine the potential value of multi-source and heterogeneous traffic data,so as to provide information with high value for the orderly development of urban traffic.Taking the analysis of multi-source heterogeneous traffic data in a city as an example,based on the deployment of Hadoop platform,the missing values of traffic flow data were supplemented by historical data repair method and time series analysis.The word frequency statistics method and k-means clustering algorithm were proposed to carry out the research on intelligent analysis technology of multi-source heterogeneous traffic data.Through the word frequency statistical method,the relationship between the weekly characteristics of traffic flow and the level of air quality is analyzed intelligently.Using time as a clustering label,through that K-means clustering method is calculated in parallel through the Hadoop cluster,the similarities and differences between different categories of traffic flow,road accidents,and vehicle illegal data were analyzed.Finally,based on visualization technology of "HTML+CSS+Java Script+Echarts",a multi-source and heterogeneous traffic big data visualization platform is designed and built to realize in-depth analysis of traffic data,providing a visualization support platform for urban traffic management,government decision-making and public travel.By using Hadoop platform to analyze multi-source heterogeneous traffic big data intelligently,not only the spatiotemporal distribution characteristics of traffic data are clarified,but also its potential value is deeply mined,which provides theoretical and technical support and effectively improves the management level of urban traffic system.Figure 38;Table 14;Reference 51...