Research On Traffic Status Discrimination Of Urban Regional Road Network Based On Multi-source Information Processing Technology

With the rapid urban growth,the scale of road network is also expanding.However,as a result of geographical location,historical culture,economic development and other factors,the structure of road network and the distribution of traffic flow in each city have their own unique features,so that different sections and roads have different influences on the traffic state of the road network,and the traffic flow operating conditions of the road network at different times and in different spaces are also different.At the same time,with the development of information technology and its extensive application in urban traffic system,massive multi-source traffic data is generated.Considering the influence factors such as static attributes,functions,levels,layout differences,and spatio-temporal imbalance of traffic flow,this thesis uses multi-source fusion data to discriminate the traffic status of urban regional road network,and obtains multi-angle and more accurate traffic operation levels,which can provide effective decision-making information for traffic planning management personnel and provide reasonable reference for travelers’ travel behaviors.Firstly,this thesis introduces the collection technology of fixed and mobile detectors in urban road network and the basic characteristics of data collection,expounds the concept system of traffic state,the selection principle and calculation method of discriminant index,and analyzes the commonly used data processing technology of discriminant index and the comparison of model advantages and disadvantages.For the data preprocessing of single fixed detector,a fault data identification algorithm based on outlier distance detection combined with threshold and traffic flow mechanism was proposed.On this basis,a data repair and completion model based on improved DE-LSTM was constructed.Aiming at the data collected by the mobile detector,Arc GIS platform was used to match the data points of the floating vehicle map.On the premise of spatio-temporal information granularity matching,a multi-source data fusion model based on improved FA-BP is constructed to improve data quality.Numerical examples show that the fault identification,repair prediction and multi-source fusion model can obtain high-quality and highly complementary traffic status information data more efficiently and accurately.According to different research levels,the urban regional road network is divided into the level of point,line and plane,and considering the different effects of different micro units on the state,the multi-attribute evaluation index system of road segment importance is established according to the static attributes of road segment,the difference of road network structure,the uneven distribution of traffic flow and the temporal and spatial correlation.Moreover,the weight optimization model based on AHP-EMW is constructed,and the importance of the road section is quantified by approaching ideal point method,which can provide a basis for the identification of multi-level traffic status of the urban road network.Finally,an effective clustering function was used to determine the state classification,and an FCM state classification model based on optimization coefficient was proposed to determine the clustering center and interval range of different class categories.In view of the fuzziness and subjectivity of traffic status classification,fuzzy evaluation was used to discriminate the traffic status of basic section units.At the level of middle-view road,intuitionistic fuzzy evaluation is used to discriminate the traffic status.Combined with the importance of the road section and the number of road vehicles,the traffic status at different times of the macro road network is weighted to obtain the traffic status at different levels of the road section,road and regional road network.The case shows that the traffic state identification model of urban regional road network based on multi-source information processing technology conforms to the objective operation rules of urban regional road network and has high accuracy.