| With the improvement of my country’s social and economic level,the travel demand of urban residents continues to increase,leading to a sharp increase in traffic travel.The existing urban road infrastructure can no longer meet the increasing travel demand,and the urban transportation system is facing a serious imbalance between supply and demand.Refinedly mastering the trajectory of vehicle travel,tracking the trajectory of vehicle travel,and providing a basis for formulating reasonable traffic control strategies are important issues facing traffic decision makers at the moment.The urban trunk line network can be regarded as the aorta of the entire city,transporting a large amount of traffic trips in the urban transportation system.Reasonably analyze the route choice behavior of vehicles in the trunk area and explore the influencing factors of vehicle travel.It is helpful to analyze the distribution of traffic flow in time and space,balance the traffic flow of the road network,and greatly improve the traffic operation efficiency of the entire road network.It is of great significance to urban construction planning and other aspects.Based on the analysis of travel choice behavior,based on RFID travel data,this article mines travel time,path familiarity,path length,as well as the proportion of expressways,the proportion of trunk roads,and the proportion of ramps,and several major influencing factors,and according to the distribution of RFID detection equipment According to the situation,the OD area is divided into the influence area of the research trunk line,and the method of obtaining the path between OD is proposed.Based on the MNL model,the above influencing factors were tested for significance,and the elastic analysis method was introduced to analyze the model results to complete the prediction of travel route selection probability.Through the BPR function,the traffic control plan is associated with the travel time,and the traffic flow of the arterial road under real-time control is allocated.The main research contents are as follows:(1)Analyze the characteristic mechanism of the travel behavior process,qualitatively analyze the impact of various influencing factors on travel route selection behavior from three aspects: personal influencing factors,travel influencing factors,and environmental influencing factors.Build and lay the theoretical foundation.(2)By analyzing the characteristics of RFID travel data,a preprocessing method of RFID data is proposed,which solves the problem of excessive RFID data volume,redundancy and lack.At the same time,the OD area is divided according to the distribution of RFID detection points,and the method for obtaining the path between OD areas is given.The travel time,path familiarity,path length,and the proportion of express roads,trunk roads,and ramps are finally determined.6 specific influencing factors,and the extraction method of each influencing factor is given.(3)Based on the theory of maximization of utility,the MNL model of vehicle travel route selection was constructed,and the relevant parameters were calibrated through the model,and the travel time,route familiarity and the proportion of express roads were determined as significant influencing factors,and the elasticity analysis method was applied to test the results.Analyze,and aim at the problem of imbalanced data volume of the model,improve the model through the sample weight method(Sample Weight).The results show that the accuracy of model prediction has increased from 92.55% to 97.87%,indicating that the method can be used in the MNL model.(4)The overall method and process of road traffic flow prediction in the area affected by arterial lines are proposed.Taking multi-lane merging control as the research object,the BPR function is introduced to correlate traffic control conditions with travel influencing factors.By changing the green letter ratio of multi-lane merging,the road traffic impedance is changed,which in turn affects the travel time of each path.The multilane control of ramp(1)of Chongqing Jinyu Interchange is selected as the research object to conduct a case study.The traffic flow parameters are obtained through field investigation,and the probability of vehicles passing the path is calculated based on the improved MNL model,and finally the road traffic flow is completed.prediction. |
