Research On Methodology Of Optimization Of Urban Traffic Network Considering Superblocks

Superblocks are city blocks whose sizes are significantly larger than average,such special structure is widespread used across cities such as China,which has attracted much attention in the field of both urban planning and traffic planning.In particular,the blocks in China are mostly closed blocks,which means that there are only a few entrances and exits connecting the traffic inside and outside the block.A few studies have proved that the network performance,such as stability,reliability,and traffic efficiency,are closely related to the road structure.However,few studies have investigated that how superblocks affect network traffic performance.In fact,there is still no consensus on which kind of street pattern can better carry traffic.On the other hand,existing researches which focused on network structure or street pattern usually convert the geography network into an abstract network by using graph theory or topology theory.In this way,the space enclosed by the roads(i.e.block)is usually ignored or "abstracted".In practical,the block is not only an important factor in describing urban space,urban layout or transportation facilities,its inside space is exactly where the traffic demand happens.Hence the form of blocks directly affects the distribution of traffic inside and outside the block.To sum up,this study aims to put forward a methodology to study the performance of networks with superblocks,further to the general road networks,by considering the traffic inside blocks.The results and other insights from this study promoting a deeper understanding of the relationship between network structure and traffic performance,and they should be of direct interest to city/traffic-planning decision-makers in dense urban cities.The research work of this thesis mainly includes the following four parts:1.Modeling the road networks by considering city block size,and evaluating the impact of block size on network traffic performance.First,this study relaxes the assumption that all demand is generated and attracted in the middle of the links or at intersections.Vehicles still enter and exit the links in their mid-points,but the actual origins and destinations are distributed uniformly within each block.In this way each trip is divided into two parts,the one that takes place inside the block,and the one outside.Second,the analytical formulations of the three indicators,i.e.travel distance,number of turning maneuvers,and travel time under low demand scenarios are constructed.Each indicator is calculated independently for inside and outside the block.An example is given to provide insights into the most important factors affecting the overall traffic performance.Last,a VISSIM simulation with a Dynamic Traffic Assignment(DTA)is used to study the traffic behaviors also for high demand scenarios.The results confirm the intuition that traffic inside the block plays an important role on the overall performance in networks with larger block size and/or short trips.In addition,smaller block size networks yield shorter travel distance and better travel accessibility,whereas larger block size networks have shorter travel time and smaller number of turning maneuvers under lower demand levels.In the congested situation,the larger block size networks show relatively better overall performance as they take longer to get congested and gridlocked.2.Modeling the road networks which contain superblocks and evaluating the impact of superblocks on network traffic performance.First,an abstract grid network with bi-directional streets is created,and a simple demand model is loaded on it to simulate a dense city environment.The traffic trips are assigned according to a Static Traffic Assignment(STA)model that leads to the user equilibrium.Both,the road impedance and the intersection delay are considered in the travel time function.Second,the node removal process is employed to generate superblocks paying attention to four different features(i.e.size,location,shape,number).Multiple scenarios reflecting changes to these four features are designed to analyze how each of these features impacts the overall network performance.The overall network performance is evaluated using travel distance,travel time,volume-to-capacity ratios on nodes and links,as well as the level of traffic heterogeneity across the network.Last,other possible scenarios,including different network sizes and other demand patterns are also discussed.The results indicate that superblocks within a certain size do not affect traffic significantly.The importance and connectivity of nodes and links surrounding large blocks are most important factors affecting the overall traffic performance.Sufficient node/link capacity and effective signal control can also help decrease the negative effects of superblocks.Furthermore,the traffic performance can be significantly improved by dividing the large block into several relatively smaller blocks.3.Proposing a methodology to design and optimize block gate setting for superblocks,and evaluating the impact of block gates on network traffic performance.First,a generalized travel cost model for two modes(i.e.private car and metro)is constructed based on the division of travel inside and outside the block.The lower level problem addresses both the mode choice and the route choice with a multimodal model describing the traffic assignment.The upper level problem minimizes the total cost,which includes the generalized costs associated with three,users related to the superblocks,users related to the background traffic and infrastructure(i.e.gates).Second,an efficient solution algorithm is then proposed to solve this bi-level optimization model.Last,a case study is used to illustrate the applicability of the model and the proposed algorithm,as well as the relationship between block gate setting and network performance.The results indicate that opening more gates is an effective way to improve network performance and reduce travelers’ cost.However,too many gates lead to marginal improvements in travel performance at a very high infrastructure cost.Moreover,the location of gates is also crucial to their impact on the overall performance.For example,it is possible to have a lower cost with 3 gates well located than with 9 gates poorly located.Hence,there is indeed an optimal number of gates to balance the benefits and costs.The optimization model proposed in this paper can effectively output the gate setting under minimum total cost,this includes also a higher mode share for the more sustainable transport mode(e.g.the metro).The cast study shows that compared with the original setting,the optimal setting can reduce the total cost by 13% and increase the sharing rate of the metro by3%4.Proposing a novel perimeter control strategy focused on networks with a closed superblock,in this way holding or evacuating the traffic flow by using superblocks.First,the urban road network is divided into three regions: a central area,a suburb area and a closed superblock.Then the traffic cost model is built by considering route choice behavior of cars.The dynamic equilibrium equations of traffic flows among three regions are established based on the Macroscopic Fundamental Diagram(MFD).Second,the perimeter control framework is built on the basis of the Model Predictive Control(MPC),in which the objective function is to maximize the vehicle completion rate of the central region,and the control variables are the tolls and number of gates at the boundaries.Last,a numerical case based on the real data is used to verify the efficient of the proposed model.The results indicates that the proposed perimeter control model can effectively alleviate traffic congestion in the central area,improve the uniform distribution of traffic density among regions,so as to the network service capacity.The cumulative number of vehicles of the central area in the end of simulation is reduced by 40% compared to the case of without the perimeter control,while the service capacity is increased by 15%.Both the tolling at perimeters and changing the number of gates of superblocks are effective way to apply the perimeter control.Tolling is used to control the traffic flow by impacting the travel behaviors,while the number of gates play an role on changing the transfer rate at perimeters.