An Urban Intersection Signal Timing Method Based On Trajectory Data

As the number of motor vehicles in cities continues to increase,urban traffic congestion is getting more and more serious.Urban traffic congestion has become an important problem that plagues residents’ travel.What’s more,as the main bottleneck restricting the efficiency of urban traffic,urban intersections’ signal control has received extensive attention.With the advent of online taxi-hailing,massive amounts of trajectory data have been generated,which greatly expanding the data sources for traffic analysis.Compared with traditional coil and bayonet data,trajectory data has the characteristics of large data volume,wide data distribution,high accuracy,and easy access.Therefore,how to apply trajectory data to intersection signal timing has important theoretical and practical significance.This thesis focuses on the intersection queuing estimation based on trajectory data and intersection signal timing methods based on queuing estimation.The main research work and innovations are as follows:1)Queue length,as an important indicator to measure the operation status of an intersection,is a significant basis for traffic signal control.In this thesis,an intersection queue length estimation method based on Bayes’ theorem is proposed,which uses the queue position of the last GPS vehicle on the lane to estimate the intersection queue length under low permeability.Simultaneously,the effect of sampling periods and permeability on this method is simulated and analyzed.Simulation results prove that this method is effective and accurate.2)Based on the derived characteristics of the trajectory data at traffic junction,the traffic wave speed estimation algorithm is proposed.Upon aggregating the parking and starting timespace points of the trajectory data of each phase into a same cycle,then the least-square method is employed to fit both the start and stop waves.Subsequently,the characteristic of the mentioned start and stop waves is further applied to estimate the intersection queue length.As such,the maximum intersection queue length and its dissipation time can be simultaneously obtained.Furthermore,our simulation results show the estimation accuracy of the proposed method,which approaches the estimation accuracy obtained by the Bayes theorem.3)An algorithm is proposed for obtaining queuing distribution at intersections based on historical queuing,which obtains the queuing distribution of stable traffic flow and unstable traffic flow.What’s more,the confidence test results of the above-mentioned two queuing estimation results for the stable traffic flow states are also analyzed.Therefore,the credibility analysis for the intersection queuing estimation results is characterized.4)According to these three methods of queue length estimation methods mentioned above,the corresponding kinematic model,shockwave model and statistical model are respectively proposed to calculate the minimum passing time for each phase of the intersection.The phase correction constant is defined to get the optimal green light duration for each phase.Through simulation verification,it is found that the timing schemes calculated by the three signal timing methods are similar.Compared with the classic Webster signal timing method and the original signal timing scheme,the simulation results show that there is a significant improvement in delay reduction,which verifies the effectiveness of the method proposed in this thesis.