Research On Data Quality Evaluation And Repair Method Of Urban Expressway Traffic Flow

In recent years,with the rapid development of China’s social economy,the urbanization process has gradually accelerated,and the urban road network structure has also become more and more perfect.In the urban road network structure,the status of the expressway is very important.The efficiency of the expressway is directly related to the operational quality of the entire urban road network,thus affecting the normal operation of the urban traffic system.In-depth mining and full use of massive traffic flow data resources can provide decision-making basis for urban traffic managers.According to the data quality evaluation of urban expressway traffic flow,taking the urban expressway microwave data as the research object,this paper uses the correlation test method to analyze the spatio-temporal characteristics of expressway traffic flow,and establishes the traffic flow data quality evaluation system.At the same time,the tensor theory is used to construct the data.The repair model provides a theoretical and methodological reference for the management and repair of expressway traffic flow data.First of all,the first chapter of this paper summarizes the current domestic and international research status of traffic flow data restoration and data quality evaluation.On this basis,the urban expressway traffic flow is determined as the research object,and the data quality is improved.The data quality evaluation and repair methods of expressway traffic flow are studied,and the technical route of this research is given.Secondly,the second chapter of this paper analyzes the characteristics of urban expressway traffic flow.The existing traffic flow data collection technology and the microwave data collection of Beijing Second Ring Expressway are introduced.On this basis,from the perspective of correlation,the time correlation of Beijing West Second Ring traffic flow data and the spatial correlation of East Second Ring traffic flow data are analyzed,and the quality evaluation of expressway traffic flow data is studied for subsequent chapters.The fix provides theoretical support.Thirdly,the third chapter of this paper establishes the evaluation system of urban expressway lane traffic flow microwave data quality.The analytic hierarchy process and fuzzy comprehensive evaluation method are used to establish the data quality evaluation system from the perspectives of "point" dimension and "line" dimension.Among them,the "point" dimension is used to describe the data quality of the single lane traffic flow in the detection section,and the "line" dimension is used to reflect the traffic flow data quality of the entire inspection section.The detection targets of the L1 lane and the East Second Ring of the H2003 detection section of the East Second Ring Road in Beijing were selected as the research objects of the "point" dimension and the "line" dimension.The fuzzy matrix was constructed by trapezoidal fuzzy numbers to establish the membership function of the evaluation index.Implement an empirical study of data quality assessment.Then,the fourth chapter of this paper describes and introduces the existing tensor theory,and expounds the tensor CP decomposition and CP-ALS algorithm.On this basis,the tensor input mode of CP-ALS is established,and the CP-ALS algorithm is compared with several other existing algorithms to test the effect of algorithm repair.Finally,the fifth chapter of this paper applies the measured traffic flow data of Beijing East 2nd Ring Expressway,and uses the proposed data quality evaluation method to achieve multiple evaluations of the "point" and "line" dimensional data quality before and after repair,and will evaluate The results were compared.The results show that the quality of the fast road traffic flow data after the tensor repair method is better than the data before the repair in the "point" and "line" dimensions.