Research On The Mehod Of Traffic Sigal Control Based On Travel Time Reliability

With the dramatic increase of motor vehicles and appearance of more and moretraffic jams, the volatility of traffic operation state within a certain time and space scopeshould not be ignored regarding to the fragile city transportation system in the currentsituation. Facing travelers’ requirements on the accuracy of travel time and trafficmanagers’ demand on meticulous management, many scholars at home and abroad haveput forward the concept of TTR (Travel Time Reliability) and its definition andcalculation model. However, past studies mainly focused on the link reliability, whichassumed that the intersection is absolutely reliable. In fact, the intersection delay is animportant factor in causing the trip travel time uncertainty.Travel time distribution of signalized intersection was studied firstly and delayestimation model and calculation method were then proposed based on TTR. Data fromintelligent monitoring and recording system of Beijing Road were used to make adetailed research on the distribution of intersection travel time by the means of asoftware named SPSS based on mathematical statistics and probability theory. In fact,the distribution pattern of intersection travel time and its delay are in consistent. Thus,the estimated model and calculation method of the intersection travel delay undercertain reliability were put forward.On the basis of above research on intersection TTR, TTR was integrated to thesignal control model. Due to the complexity of the model optimization target, geneticalgorithm was applied to solve this model. The paper took an intersection in Beijing asexample to analyze the difference between models based on different targets. It showedthe result from the model in this paper existed significant difference in contrary toclassic models in the case of large flow, but kept consistence in the case of low flow.After the research on single point control, the vehicle’s unblocked rate was used todefine the arterial TTR, which was taken as optimization target in the new model,considering the characteristic of arterial traffic signal coordination control. Then, anexample was designed and solved by genetic algorithm. Results showed that thecommon cycle length of the new model was longer than other models; the phase offsetof the new model is basically in consistence with models based on the maximum of the green wave but is different from models based on the minimum of delay.Travel path is made up of road section and intersection, so travel time reliability isjointly decided by these two aspects. In this study, travel time reliability for a singleintersection of signal point control and for multiple intersections of arterial coordinationcontrol was defined first and, further more, new signal control models based on TTRand solve method for models were put forward.