Literature Review of Macroscopic Fundamental Diagram

Name: Peng Jixian Student NO.: 3030016479

MFD (Macroscopic Fundamental Diagram) of traffic flow refers to the relationship between average flow and average traffic density in a network. Since Wardrop (1968) proposed the monotonically decreasing relation for average speed and flow, in the last few decades, many experiments and simulations have been carried out in this field. Literature about existence and properties of MFDs of a single highway is extensive. Until recently, researchers found proof for existence of MFDs on a whole neighborhood. Geroliminis and Daganzo (2008) held a field experiment in Yokohama (Japan) based on data from fixed detectors and floating vehicle probes. The study shows that the curves of MFD exist in a large urban area allowing for experimental error. Furthermore, they confirmed the existence of MFD on any street with no turns by giving exact analysis of a shortest path recipe for street capacity and approximate formulas. Geroliminis (2008) presented a finding of that MFD relating production and accumulation exists for 5-10 neighborhoods, independently of their OD

tables, and also, rush hour can be predicted by the fix relation between space-mean flows and trip completion rates. Therefore, in later articles the term MFD is also used to represent the relation between network outflow and accumulation, many investigations have been made to study the factors influencing the shape and properties of MFD. In the early stage, investigations concentrated on the evidence for MFDs existence on a network; later on, researchers began to explore the properties of MFDs of the neighborhood. The most common phenomena is hysteresis property of MFDs. Geroliminis and Suns (2011) research also shows that for the same average network density, network flows are higher in the onset and lower in the offset of congestion. It is possible to forecast outflow given only its network density in a dynamic system with no hysteresis and a well-defined MFD, while not possible in a system with hysteresis
1

unless the systems current state is known. Experimental studies show that MFD can be well predicted for networks consisting of a single route, while the congestion flows observed in networks consisting of multiple overlapping routes for the same density are less. Carlos F. Daganzo.et al (2011) explained this phenomenon by pointing out the instability of flow distribution when the network density gets sufficiently high. They also discovered a specific critical density at which small idealized networks undergo a bifurcation of the order of MFDs. Furthermore, Daganzo (2011) analyzed the stability of flow and density distribution of freeway traffic; he discovered that unless the freeway is either uncongested or congested with average density increase with time, such distribution tends towards uneven. He also quantified the effects of this instability with formulas and verified with simulations. Geroliminis and Sun (2011) investigated the effect of spatial distribution of vehicle density on the scatter and shape of MFDs. Well defined MFDs have been proved to exist on homogeneous networks. Network characteristics influence the shape of MFDs, though some of them have been identified, further exploration is required, ex. the geographical configuration of the network, which influence vehicle velocities and driving behaviors. Previous researches studying MFDs have been mostly based on loop detector data or GPS data, while it is difficult to change traffic OD matrices in real world if we want to investigate MFDs properties in different network conditions. Therefore, much effort has been devoted to modeling and simulation for MFDs description. Taylor et al (2004) presented a Statistical Traffic Model based on MFD and conservation of vehicles using Matlab/Simulink and applied the model to help design a ramp metering algorithm. Courbon and Leclercq (2011) simulated an equilibrium model to determine traffic situations and examined three common used methods to estimate MFD. They also investigated that how heterogeneities in traffic states influence MFD, and raised recommendations to better calibrate them. However, researches also show that sometimes MFDs cannot be perfectly applied for real situation prediction. Liu et al research (2011) shows that supply curve generated from MFD can not accurately reflect the real origin-destination movements in
2

networks for higher levels of demand and they proposed these curves to be approximated from micro simulation under congested conditions. Ji et.al (2010) built a microscopic simulation model in Vissim to derive and investigate the shape of MFDs by changing traffic conditions of different types of network. They found metering control of freeways and changing traffic demand affect directly on the shape of MFDs. However, their work was based on a single highway model, not a network. The basic proof for MFDs existence in neighborhood has been accomplished by pioneering researchers, though, further explorations are required to investigate the complicated relationship between MFD elements, e.g. to examine the properties of MFDs properties in dynamic traffic assignment or to applied the network model into a real word neighborhood to study its performance in different traffic conditions.

References Bonzani, I., & Mussone, L. (2009). On the derivation of the velocity and fundamental traffic flow diagram from the modelling of the vehicledriver behaviors. Mathematical and Computer Modelling, 50(7-8), 1107-1112. doi: 10.1016/j.mcm.2009.06.004 Courbon, T., & Leclercq, L. (2011). Cross-comparison of Macroscopic Fundamental Diagram Estimation Methods. Procedia - Social and Behavioral Sciences, 20, 417-426. doi: 10.1016/j.sbspro.2011.08.048 Daganzo, C., & Geroliminis, N. (2008). An analytical approximation for the macroscopic fundamental diagram of urban traffic. Transportation Research Part B: Methodological, 42(9), 771-781. doi: 10.1016/j.trb.2008.06.008 Daganzo, C. F. (2011). On the macroscopic stability of freeway traffic. Transportation Research Part B: Methodological, 45(5), 782-788. doi: 10.1016/j.trb.2011.02.001 Daganzo, C. F., Gayah, V. V., & Gonzales, E. J. (2011). Macroscopic relations of urban traffic variables: Bifurcations, multivaluedness and instability. Transportation Research Part B: Methodological, 45(1), 278-288. doi: 10.1016/j.trb.2010.06.006 Farhi, N., Goursat, M., & Quadrat, J. P. (2011). The traffic phases of road networks. Transportation Research Part C: Emerging Technologies, 19(1), 85-102. doi: 10.1016/j.trc.2010.03.011 Gayah, V. V., & Daganzo, C. F. (2011). Clockwise hysteresis loops in the Macroscopic Fundamental Diagram: An effect of network instability. Transportation Research Part B: Methodological, 45(4), 643-655. doi: 10.1016/j.trb.2010.11.006 Geroliminis, N., & Daganzo, C. (2008). Existence of urban-scale macroscopic fundamental diagrams: Some experimental findings. Transportation Research Part B: Methodological, 42(9), 759-770. doi: 10.1016/j.trb.2008.02.002 Geroliminis, N., & Sun, J. (2011). Hysteresis phenomena of a Macroscopic Fundamental Diagram in freeway networks. Transportation Research Part A: Policy and Practice, 45(9), 966-979. doi: 10.1016/j.tra.2011.04.004 Geroliminis, N., & Sun, J. (2011). Properties of a well-defined macroscopic fundamental diagram for urban traffic. Transportation Research Part B: Methodological, 45(3), 605-617. doi: 10.1016/j.trb.2010.11.004
4

James Taylor, C., McKenna, P. G., Young, P. C., Chotai, A., & Mackinnon, M. (2004). Macroscopic traffic flow modelling and ramp metering control using Matlab/Simulink. Environmental Modelling & Software, 19(10), 975-988. doi: 10.1016/j.envsoft.2003.11.009

Liu, R., May, T., & Shepherd, S. (2011). On the Fundamental Diagram and Supply Curves for Congested Urban Networks. Procedia - Social and Behavioral Sciences, 17, 229-246. doi: 10.1016/j.sbspro.2011.04.516 Qian, X., Hoogendoorn-Lanser, S., Hoogendoorn, S., Daamen, W., & Ji, Y. (2010). Investigating the Shape of the Macroscopic Fundamental Diagram Using Simulation Data. Transportation Research Record: Journal of the Transportation Research Board, 2161(-1), 40-48. doi: 10.3141/2161-05

Homework of thesis writing introduction Student name: Peng Jixian Student number: 3030016479 Critique of a PHD thesis Thesis title: Dynamic Macroscopic Modeling of Highway Traffic Flows Author: Wong Chun Kuen Wong proposed a Multi-Class Lighthill-Witham-Richards (MCLWR) model in his PHD thesis. Compared to a Single-Class Lighthill-Witham-Richards (SCLWR) model, MCLWR model can better represent real world traffic situation. He also raised numerical examples to demonstrate that the MCLWR model can simulate three main puzzling phenomena of traffic flow: two-capacity phenomenon, hysteresis phenomenon and platoon dispersion. In the last part of main body, he developed a fifth-order WENO scheme, which showed higher efficiency than the Lax-Friedrichs finite difference scheme, to solve the MCLWR model. In terms of topic selection and approach design, the author has made sound choices. Traffic flow theory investigates traffic phenomena resulted from complex interactions of vehicles, and flow modeling plays a significant role to help analyze characteristics of traffic flow in different conditions. Wongs thesis focused on the LWR model, the simplest and most classical traffic flow model which has stood the test of time and kept attracting attention and new explorations. Wong limited the study to the extension from SCLWR model to MCLWR model, but he has dogged the question deep enough to accomplish a valuable research. The WENO scheme has not been applied to solve traffic flow models until recently, and in this study, it proved to be accurate and reliable for partial differential equations of the MCLWR model. Wong has also developed an algorithm to solve MCLWR model using the existing Lax-Friedrichs finite difference scheme to compare with the WENO scheme, though the former can only be applied for MCLWR models of low orders, the comparison has made the argument more convincing and completed. The literature review was organized in a clear structure and logical order making it easy to read. The reference resources had been well selected and the development of LWR model exploration was thoroughly described. Then the author presented the
6

three puzzling phenomena, which he used the MCLWR model to predict in the following chapters, in the second part of literature review with a distinct transition paragraph. However, review of literature about WENO scheme which was an important result of the whole study, should be included in the end of this chapter. Overall, this is a comprehensive and in-depth study of traffic flow model. The author firstly described SCLWR model and its basic properties, deepened the analysis into shock fitting algorithm, and then expand the algorithm to MCLWR model. In both models presentation, he utilized numerical examples to show their performance. Nevertheless, the content would have been more completed if the author had finished the model calibration and validation basing on empirical traffic data. Meanwhile, other properties of MCLWR model apart from the three puzzling phenomena should have been included in the model analysis. In the concluding chapter, Wong has well examined his own study and pertinently proposed a few directions for further research, such as to classify the driver classes in other respects and to fully explore the MCLWR model through field observation, therefore the demonstration has been enhanced. The whole thesis was organized in a typical traditional-simple structure. In the main body, each chapter was started with an introduction and ended with a concluding remark, this helps readers to grasp the main ideas and follow the authors train of thought easily. In chapter 4, 5 and 6, macroscopic fundamental diagram were presented to illustrate the models performance. The layout was appropriate but discussion paragraphs were missing for each plot figure. Tiny mistakes exist in the table of contents: the title of chapter 5 should be analytical shock fitting algorithm for MCLWR model, not analytical shock fitting algorithm for SCLWR model; 4.8 Closure did appear in the thesis but is missing in the table of contents. Since many basic properties are shared by MCLWR and SCLWR model, the author used an independent chapter to demonstrate shock fitting algorithm for SCLWR model instead of adding it into the introductory chapter of SCLWR model, this makes
7

the presentation cohesively conveyed. Taken as a whole, the writing style in this thesis is concise and apropos. The author did not use many hedges; the research methodology and computing results were clearly and continuously illustrated; and also, the comments and critique in the literature review had been made appropriately and in good balance. However, it is interesting to find that the author used so many it is interesting to find that in the thesis, and the paragraphs in the conclusion chapter were almost duplicated from the introduction chapter. In conclusion, Wong has made a meritorious study and shown a broad knowledge of traffic flow modeling and excellent writing skills in his PHD thesis.