Road-net Condition Information Theory And Methods In Internet Of Vehicle Circumstances

With the increasing tendency of vehicle ownership, road traffic congestion, traffic accidents and resulting pollution are difficult problems need to be solved at home and abroad. At present, the main approaches to solve the above problems are to increase share of urban roads, restrictions, and to apply intelligent transportation technology. Among those approaches, the application of intelligent transportation technology is an internationally recognized fundamental way to solve the traffic problem, but the lack of comprehensiveness and accuracy of the information are the main reasons to make intelligent transportation technology impractical.In the thesis, I developed systematic physical and Internet architecture of telematics, and I also proposed the main technologies about situation information awareness of network traffic network under the environment of telematics,Because there are large volumes of anomalies in the original data of telematics RFID sensors, I considered the three traffic flow parameters as a whole. On the basis of establishing data sampling interval and the quantitative relationship of data dispersion level models, I presented a four-step screening method for outliers, and four data recovery methods on different occasions, and then establishing a standardized data filtering and recovery processes. The applicability and validity of the method are verified by the actual data of Changchun City, and it has higher accuracy and stronger engineering practical value.In the thesis, I also proposed a road travel time estimation model, which only relies on GPS data in internet of vehicle. This mode solve the problem of road travel time estimation based on coarse-grained(30s or so) floating car data effectively, with no intersection signal timing parameters. To Qinhuangdao main road test vehicle field with vehicle test, this model the model of travel time estimation values of mean absolute error and the mean relative error are 12 s and 8.67%, due to the traditional method.Based on the analysis on the characteristics of road traffic parameters after traffic accidents, I presented an automatic recognition method for traffic accidents, which combine the collecting telematics flow data and travel time estimation data. Through the establishment of the evaluation index, the simulation software of VISSIM is used to show that the new method has the efficiency and whether it adapts to different traffic conditions.This article is mainly for awareness field of telematices, and provides the basic data support for telematics application.