Study On Intesection Signal Control Based On Queue Equalization And Minimum Delay

With the social and economic development,and the rapid increase in car ownership,the problem of urban traffic congestion is becoming increasingly serious.Intersections are important nodes in the urban transportation network.Traffic congestion problems are more prominent at intersections.Some intersections even have queue overflow phenomenon,which seriously affects the normal operation of adjacent intersections,and leads to the gradual expansion of traffic congestion.At present,the way to deal with this kind of traffic problem is that the traffic police conduct traffic command at the intersection,which costs a lot of manpower and financial resources.Therefore,it is of practical significance and value to study the queue overflow problem at urban signal intersection.In order to make full use of the green time at signalized intersections,avoid queue overflow at intersections,and reduce the average vehicle delay,this paper takes queue equalization and minimum delay as the optimization objectives to control signal intersections dynamically.Specific is as follows:(1)Generally,the error of estimation value of operation state at intersections based on a single traffic data is large.Therefore,the characteristics of different traffic data are summarized from the aspects of data collection technology and detection equipment,and the moving data and fixed point data are selected for the state estimation of the intersections.The processing methods of different data were analyzed.According to the GPS data of vehicles,the queue length of each cycle at the intersection was estimated by the sum of the floating vehicle position at the tail of the queue and the queue length of the remaining vehicles.Based on the traffic wave principle,the queue length of vehicles is estimated based on the section detection data.After analyzing the estimation results of two kinds of data,a method of queue length estimation based on GPS and section detection data is proposed.(2)In order to avoid unbalanced phase queue length,the phenomenon that vehicles cannot dissipate as soon as possible and queue overflow occurs,more green time is allocated to the phase with longer queue length.In order to reduce the average delay of vehicles,three delay models(steady state equilibrium,constant number and transition function delay model)are analyzed,and a signal control model based on queue equalization and minimum delay is proposed.With the method of fuzzy compromise planning,queue equalization and minimum delay is the simplified as a single optimization goal.The principle of particle swarm optimization is analyzed and the signal control model is solved.(3)Taoyuan Street-Taohua Road intersection in Nanchang city was simulated to obtain GPS data and section detection data,and the queue length estimation methods based on GPS data,section detection and two types data were verified respectively.Get the mean absolute relative error of each method are 11.27,14.71 and 10.40 respectively.Thus the accuracy of queue length estimation based on two types data was significantly better than that based on single data.Verify the signal control model,the dynamic signal control is carried out during the peak period of the intersection.Compared with the current situation,the average vehicle delay at the intersection was reduced by 8.26 s and the average queue length was reduced by 3.25 m.Compared with Webster timing method,the average delay of vehicles at intersections was reduced by 3.76 s and the average queue length was reduced by 1.01 m.It can be seen that the signal control model based on queue equalization and minimum delay can effectively reduce the queue length and average vehicle delay at intersections,and improve the queue unbalance at intersections.