The Study Of Stochastic Model Of Left-turn Traffic Flow At Signalized Intersections

As key nodes of urban roads,signalized intersections are prone to traffic conflicts,blockages,and vehicle emissions problems.Therefore,scientific evaluation of the traffic state at signalized intersections is fundamental for modern traffic optimization control.The original route could be improved by including a left-turn bay by taking into account the effectiveness and safety of urban traffic.The advantage of this approach is that it can improve the traffic efficiency and safety of vehicles at signalized intersections without the need for large-scale road reconstruction.The disadvantage is that if the design of the signalized intersection is not perfect,it will lead to overflow and blockage of the queuing vehicles,which will eventually result in a waste of time and space resources and additional vehicle emissions at the intersection.At the same time,relying solely on metrics such as average delay and average queue length is not sufficient to assess the stochasticity of the formation and dissipation processes of vehicle queues at such intersections.Therefore,this thesis implements a quantitative description of the formation and dissipation of left-turn traffic flow queues under different left-turn control strategies through the analytical approach of stochastic queueing models.On the basis of this,a thorough framework for traffic performance analysis is created in order to clarify the effects of traffic demand,the length of the left-turn bay,and signal timing on vehicle queuing.This framework will provide a theoretical basis for operational state prediction and traffic performance evaluation of signalized intersections with a left-turn bay,as well as for further optimization of the design of such intersections.The following is a list of this thesis’ principal work:First,we develop a stochastic queueing model for vehicles at intersections with a left-turn bay by focusing on the overflow and blockage problems that are caused by the addition of a left-turn bay under the protection of the left-turn signal control strategy.The steady-state probability distribution of the queue length of vehicles in the subject lane at an arbitrary time is obtained using matrix analysis techniques.On this basis,we obtain the basic average performance index of vehicle queueing at intersections and the signal-state-dependent queue length and its variance,which can quantify the dynamics of vehicle queueing at intersections.Based on the validation of the model,numerical experiments were conducted to clarify the factors that affect the operational efficiency of the signalized intersection with a left-turn bay.Second,to address the problem that protected-only phasing leads to a certain amount of temporal and spatial waste of intersection resources,we develop a stochastic queueing model for vehicles at a signalized intersection with a left-turn bay under the protected/permitted signal control strategy.Compared to the previous model,this model has a wider range of applications and can be converted to the previous model under certain conditions.By solving this model,it is possible to obtain the average performance index of vehicles queuing at intersections and the signal-state-dependent queue length and its variance.This model can be used to analyze the queueing behavior of vehicles at signalized intersections with a left-turn bay under permitted or protected/permitted phasing.On the basis of verifying the validity of the model,the influencing factors of intersection capacity under the protection/permission signal control strategy are further analyzed through numerical experiments.Finally,in order to solve the problem of additional delays and emissions caused by increasing the left-turn bay,the number of stops of vehicles at intersections is obtained based on the established random queuing model,and the vehicle emissions at intersections are quantified from a microscopic perspective by the vehicle specific power method(VSP).The analysis of vehicle emissions at a signalized intersection with a left-turn bay was performed using numerical experiments under various factor variations.As a result,in this thesis,we construct a comprehensive quantitative analysis framework that can evaluate the capacity,vehicle queuing dynamics,and vehicle emissions of signalized intersections with a left-turn bay,which can provide methodological support for the application of intersections with a leftturn bay.