Research On Traffic Assignment Of Urban Traffic Network Based On Travel Time Analysis

Because of the influences of all kinds of stochastic events on actual traffic network,traffic supply and demand are both uncertain,and this uncertainty leads to the stochastic change of travle time.One of the research point is how to predict travle time instantly and accurately.So the travel time prediction with temporal and spatial covariance effects are investigated in urban network.Travelers with different risk preference will choose different path in order to arrive at their destinations fast,smoothly,and reliably.It is necessary to identify travelers' risk preference and study their route choice behaviors.Then the road traffic assignment problem under stochastic circumstances and network design problem with lane reallocation are also studied.On the basis of the existing results,some studies have been carried out and the main contributions of this thesis are shown as follows:1.Temporal-spatial correlation effects on vehicle travel time prediction.It is recognized that travel times on a link are temporally correlated with its travel times of previous time periods.Also,the link travel time are spatially correlated by travel times on its neighboring links.Based on such temporal and spatial correlations,a new nonparametric method,employing the temporal and spatial variance-covariance of travel times on related links together with historical travel time data,is proposed for travel time prediction in urban roads.The proposed method is capable of rapidly predicting the link travel time for different time periods in the near future.In order to ensure travelers arrive at destination reliably,the uncertainty of travel time should be considered.Observation surveys are used to validate the travel times predicted by the proposed method.Compared with some other existing travel time prediction methods on the basis of the same database,the proposed method is proved to be fast and accurate.2.Modeling travelers' behavior in route choice based on risk preference.The classical indexes used in travel time reliability are limited in demonstrating the risk aversion of different travelers.It is discovered that the reliable route choice model based on mean and standard deviation are all established on a same selected route set essentially.That is to say,without consideration of travelers' risk preference,only a part of all the paths is rational travelers alternative reference which forms a set called effective path set.Once the distributions of the paths in traffic network are defined,the effective path set could also be identified with a given traffic assignment.The most satisfied paths from the effective path set,which can be called the optimal paths,will be selected by travelers in different risk preference style.Based on the expected utility theory and the travel time budget,a path selection method with the mean and standard deviation of travel time is proposed.This method can provide the more reliable path to each traveler with different risk preference.3.A fixed-demand user equilibrium model is established with consideration of travel time reliability.According to the research of route choice under stochastic travel time in previous part,it is found that the paths with the minimum certainty equivalent of disutility to travelers will be selected all the time.Then the traffic assignment with these behaviors of route choice comes into a new user equilibrium.On the other hand,travelers will set expectations up front and these expectations may also affect the user equilibrium assignment.So the expectations of travelers before their trips should be set reasonably.The different user equilibrium assignment with different expectations should be researched.And a novel algorithm is proposed to solve the model.4.Network design modle with lane reallocation and certainty equivalent of disutility is established.Within the morning and evening rush hour,the two-way road flows are always unbalanced in opposite directions.In order to make full advantage of the existing lanes,the two-way road lane has to be reallocated to play the best role in managing congestion.On the other hand,travelers hope to reach the destinations on time.So as to alleviate the commute congestion in urban transportation network and improve the utilization of each lane,a bilevel programming model is established with fixed demand.In the bilevel model,the government at the upper level reallocates lanes on the two-way road to minimize the total system certainty equivalent of disutility.The traveler at the lower level chooses the optimal route on the basis of his certainty equivalent of disutility of the path involved.