Data Analysis System For Automotive Electronic Indentification

With the rapid development of the national economy,the number of motor vehicles on the road rises continuously,which has increased the pressure on road traffic and caused traffic congestion.This has improved travel cost and caused many problems such as environment pollution,which has increased the operating costs of society.Radio frequency identification(RFID)has been widely used in the field of transportation due to its advantages such as high recognition accuracy is not interfered by weather and other factors,identified quickly and containing more comprehensive information.Forming electronic registration identification of the motor vehicle provides a new means to solve urban road traffic problems.Installing automotive electronic identification tags on motor vehicles and automotive electronic identification equipment above the collection point road collect data,process the collected traffic data,analyze the behavior of vehicles,detect accompanying vehicles,identify road traffic congestion and predict traffic flow.This paper designs a data analysis system for automotive electronic identification based on Shenzhen automotive electronic identification data.The travel characteristics of urban roads were analyzed with the help of traffic flow parameters,and the accompanying behaviors of vehicles,traffic congestion detection,and short-term traffic flow prediction were analyzed based on vehicle electronic identification data in a period of time.It provides reasonable suggestions for road traffic management departments to alleviate urban road traffic problems.The main work of this thesis are as follows.Accompanying vehicle detection.Accompanying vehicle is a concept in the field of transportation,which means that the number of certain vehicles passing through the reading site reaches a certain threshold in a certain period of time,and the accompanying vehicle detection is mainly aimed at vehicles that may cooperate in the crime.The accompanying vehicle detection in this thesis mainly uses the car’s electronic identification obtained from various collection points.Detect accompanying vehicle by using the FP-growth algorithm which based on the Spark platform,processing of automotive electronic identification data using key-value pairs,the efficiency of accompanying vehicle detection is improved.Traffic congestion detection.On the basis of vehicle electronic identification data,the traffic volume and speed are selected as the main traffic flow parameters to determine the traffic congestion status of the road.The FCM algorithm with fuzzy theory,according to the actual traffic congestion status of different roads,for the difference of clustering centers of different sample data,use the method of moving average to determine the initial clustering centers.At the same time,the values of fuzzy index and number of clusters are analyzed to obtain the best clustering effect,to obtain the recognition results of the road segment status,and to analyze the results of the road segment congestion discrimination,using model evaluation index,verifying the accuracy of the algorithm and the improved effectiveness through parallelism.Short-term traffic flow prediction.According to traffic flow data has the characteristics of time series and periodicity,a short-term traffic flow prediction model based on ARIMA was established.The use of time series ARIMA algorithm in short-term traffic flow prediction can achieve better results,but the ARIMA model can only analyze the linear nature of the data,and there will be some errors for unstable peak fluctuations.The prediction error at the peak is relatively large.Due to the complexity of the actual traffic flow,it is difficult to obtain a good prediction using only the ARIMA model.This paper proposes the ARIMA-GARCH model for short-term traffic flow prediction,which solves the non-linear characteristics of traffic flow with an accuracy rate of 94% and use the Spark platform to improve execution efficiency.Above the model research Foundation,a data analysis system based on automotive electronic identification is designed and implemented to store,process and analyze automotive electronic identification data based on the Spark platform and Kafka message middleware;The outline design of system and the architecture of system,the design of database,as well as the detailed design of data management module,accompanying vehicle detection module,traffic congestion discrimination module,and short-term traffic flow prediction module are given.