Research On Traffic Operation Risk Prediction Method Of Urban Bridge-Tunnel Joint Section Based On Multi-source Data

With the continuous increase of urban road construction mileage and traffic demand,due to urban terrain restrictions,elevated,cross-river bridges,mountain tunnels,and tunnels under the tunnel have gradually become an important part of urban threedimensional traffic.The connection between bridges and tunnels in cities can also be seen everywhere.It is inevitable that some bridges will be combined with tunnels,especially in Chongqing,which is known as the “mountain city”.Bridges and tunnels occupy a relatively high proportion of urban roads.In urban roads,the special structures of bridges and tunnels are more difficult than the traffic environment of ordinary urban roads,especially the combination of bridges and tunnels,which are affected by bridges and tunnels,and the various intersecting forms of the combined flow zones make the sections The operation situation is more complicated and has become the main node of urban road congestion.Therefore,in order to improve the smoothness of urban road traffic,it is necessary to study the traffic operation risk prediction of the urban bridge and tunnel combined section.Based on the smooth operation of traffic,the definition and classification of urban bridge and tunnel joints are defined by the safe distance.The combined segment definition model is proposed to study the complex traffic environment and traffic flow characteristics.Based on this,a variety of data acquisition techniques are studied.Based on the measured data,the traffic anomaly data is eliminated based on the threshold theory and traffic flow mechanism theory.The noise data is denoised based on Savitzky-Golay and adaptive weighted average method.Processing and multi-source data fusion;The support vector machine insensitive coefficient,penalty coefficient,kernel function and its parameter optimization selection are studied.Combined with the characteristics of traffic flow data,this paper proposes to construct a short-term traffic flow prediction model based on GA-SVR by using the powerful global search ability of genetic algorithm.The support vector machine performs parameter optimization,and uses the training set to solve its parameters to realize short-term prediction of traffic flow;Based on the prediction data,a fuzzy C-means clustering model based on RelirfF weighting is established to judge the traffic operation risk level,and the traffic operation risk shortterm prediction is realized.Finally,based on the measured data of Chongqing City,the traffic development of the main city and the operation of the bridge and tunnel are introduced.On the basis of this,the combination of the Huanghuayuan Bridge and the Shihuang Tunnel is taken as an example.The simulation results show that the research results are scientific,effective and feasible.Through the research,the thesis comprehensively analyzes the traffic environment and traffic flow characteristics of the urban bridge and tunnel combined section.In the multi-source data processing,the multi-source data fusion theory based on adaptive weighted average is proposed.A GA-SVR model based on genetic algorithm optimization is proposed for short-term traffic flow prediction,and a weighted fuzzy C-means clustering method is proposed to discriminate the prediction data.Provide theoretical basis for the study of traffic operation of urban bridge and tunnel junctions.But the traffic operation system is an extremely complex dynamic system with the characteristics of large random interference and strong uncertainty,which can be further studied in future research.