Research On Vehicle Traffic Flow Model At Signalized Intersection

Intersection is the bottleneck of road traffic.The orderly operation of intersections is of great significance to improve the efficiency of urban traffic operation.Micro traffic simulation is an effective basis for intersection design,management and control scheme evaluation.However,the existing micro traffic simulation is mainly applied to the traffic with obvious Lane settings.There are few studies on intersection simulation.Aiming at the weakening of the concept of intersection lane and the discreteness of traffic trajectory,this study establishes a conflict traffic flow model that can describe the discreteness of intersection traffic trajectory.The model can describe the real driver behavior and accurately describe the discrete characteristics of intersection traffic trajectory.Firstly,the traffic trajectory model is established.According to the weakening of the concept of intersection lane,a two-dimensional intersection traffic trajectory model which can describe the driver’s behavior is established.The model takes the initial state and terminal state as inputs,takes the minimum expected total cost of the driver as the optimization objective,and generates the optimal trajectory based on the optimal control theory.At the same time,the description,rationality and accuracy of the model are verified.Secondly,a parameter fitting method suitable for traffic trajectory model is proposed.Based on the original multivariable distribution parameter fitting method,the idea of clustering is introduced,and a multivariable distribution method based on clustering is proposed.First classify the drivers and then fit them.Through the actual data simulation,it is verified that the multivariable distribution method based on clustering can well describe the actual situation of intersection traffic flow,and has strong adaptability.For the intersection involved in calibration,the coincidence degree of trajectory distribution can reach more than 90%.For the new intersections not involved in the calibration,the coincidence degree of trajectory distribution can also reach 88.5%.Finally,the conflict traffic flow model is established.Based on the traffic trajectory model,considering the different situations of straight traffic flow and opposite left turn traffic flow at the intersection,the models are established respectively,and the simulation framework is established.The minimum relative error between the actual trajectory SD and the simulated trajectory SD is 0.15%.The maximum relative error is16.93%.The average relative error is 8.92%.Finally,the effects of intersection space size,intersection angle and flow on trajectory dispersion are discussed.It is found that the larger the intersection is,the greater the trajectory dispersion is.The greater the intersection angle,the smaller the trajectory dispersion.The greater the intersection flow,the greater the trajectory dispersion.The traffic flow model established in this study complements the deficiency of micro traffic simulation in the field of intersection,and provides a theoretical basis and evaluation method for intersection design,control and management.