This document presents a queueing based traffic flow model to model traffic behavior analytically using queuing theory. It develops the M/M/1, M/G/1, and G/G/1 queueing models and uses them to construct speed-flow-density diagrams. It then applies the models to traffic count data from a highway in New Delhi to illustrate the results. The models allow experiments to help with traffic management, congestion control, and assessing environmental impacts of traffic flows.
Download as PPTX, PDF, TXT or read online on Scribd
0 ratings0% found this document useful (0 votes)
47 views
A Queueing Based Traffic Flow Model
This document presents a queueing based traffic flow model to model traffic behavior analytically using queuing theory. It develops the M/M/1, M/G/1, and G/G/1 queueing models and uses them to construct speed-flow-density diagrams. It then applies the models to traffic count data from a highway in New Delhi to illustrate the results. The models allow experiments to help with traffic management, congestion control, and assessing environmental impacts of traffic flows.
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 23
Aditya Dhagat
-Shivam Asthana -Tanuj Verma -Rishabh Sawhney
A QUEUEING BASED TRAFFIC FLOW
MODEL INTRODUCTION The assessment of uninterrupted traffic flow is traditionally based on empirical methods. We develop some analytic queuing models based on traffic counts and we model the behavior of traffic flows as a function of some of the most relevant determinants. These analytic models allow for experiments, which pave the way towards our research objectives like traffic management, congestion control, traffic design and the environmental impact of road traffic (e.g. emission models). We illustrate our results for a highway, based on counted traffic flows. When modeling the environmental impact of road traffic, we can distinguish between both a static and dynamic impact of infrastructures and vehicles on emissions and waste . In addition, they may cause damage to natural watercourses or threaten the natural habitat of wildlife. Vehicles in turn consume natural resources and impose a strain on the environment at the end of their life cycle.
Furthermore, an increase in garbage, accidents (physical and
material damage) and, occasionally, distortion of infrastructures and nature elements (trees, animals, etc.) can be observed. Because traffic flows are a function of both the number of vehicles on the roads and the vehicle speed, the resulting environmental impact will also be a function of these parameters. QUEUING THEORY As traffic flows occupy a central position in the assessment of road traffic emissions, a robust traffic flow model is required. Traditionally, traffic flows are modeled empirically, using origin destination matrices. Queueing theory is almost exclusively used to describe traffic behavior at signalized and unsignalized intersections. The speed-flow-density diagrams will be constructed analytically, which makes it possible to assess the (future) environmental impact, perform sensitivity analyses and facilitate congestion management. VARIOUS PARAMETERS USED In our model we define C as the maximum traffic density. Roads are divided into segments of equal length 1/C, which matches the minimal length needed by one vehicle on that particular road. Each road segment is considered as a service station, in which vehicles arrive at rate and get served at rate . We define W as the total time a vehicle spends in the system, which equals the sum of waiting time (due to congestion) and service time. The higher the traffic intensity, the higher the time in the system becomes (the exact relation between W and depends upon the specific queuing model). The relative speed r, by definition, becomes: Plotting the traffic flow, density and (relative and effective) speed on a graph gives us the well known speed-flow-density diagrams. The exact shape of these diagrams depends upon the queueing model and the characteristics of the arrival and service processes. In the sequel of this paper, we will develop the following queueing models with their respective underlying CHARACTERISTICS OF THE USED QUEUING MODELS THE M/M/1 MODEL The inter-arrival times are exponentially distributed (the arrival rate follows a Poisson distribution) with expected inter-arrival time equal to 1/ (with equal to the product of the traffic density E and the nominal speed SN). The service time delineates the time needed for a vehicle to pass one road segment and is exponentially distributed with expected service time (the service rate follows a Poisson distribution). When a vehicle drives at nominal speed SN, service time can be written as: 1 / (SN * C) and equals the product of nominal speed SN with the maximum traffic density C. THE M/G/1 MODEL As in the M/M/1 model inter-arrival times follow an exponential distribution with expected inter-arrival time 1/, being the product of traffic density and nominal speed.
The service time however is generally distributed with an
expected service time of 1/ and a standard deviation of . Expected service rate is , which equals the product of nominal speed SN with maximum traffic density C.
Combining Littles theorem and the Pollaczek-Khintchine
formula for L5 (Hillier and Liebermann, 1995) and substituting for and , we obtain the following formula for the total time in the system W: Using these formulas we can construct the speed-flow, speed-density and flow-density diagram for the M/G/1 model. The exact shape of these curves depends upon the variation coefficient of the service time, . THE G/G/1 MODEL With the G/G/1 model both arrival times and service times follow a general distribution with expected arrival time 1/ and standard deviation a, expected service times 1/ and standard deviation of b respectively. Consequently, the shape of the speed-flow-density diagrams will depend not only on the variance of the service times but also on the variance of the inter-arrival times. Combining Littles theorem and the Kraemer-Lagenbach-Belz (Kraemer and Lagenbach-Belz, 1976) formula for L and substituting for and , we obtain the following formulas for the total time in the system W: APPLICATIONS/PROBLEM This application shows some results of the G/G/1 model and the state dependent G/G/1 model for highway. In this example we make the nominal speed SN equal to the maximum speed allowed on a highway in New Delhi (or 120 km/h). We dispose of hourly traffic counts Traffic flow peaks at 8.00 a.m. with 3200 vehicles per hour and then remains fairly constant at approximately 2000 vehicles per hour until 7.00 p.m. OBSERVATIONS
The maximum traffic flow observed in the three
years considered is 3350 vehicles per hour. Assuming that this flow occurred in the best possible conditions (e.g. good weather) , we calculate the maximum traffic density C with qmax equal to 3350 and the variation coefficients both equal to 0.5 (low variance). C is then equal to 74 vehicles per kilometer. CONCLUSION Based on queueing theory we analytically constructed the well-known speed-flow-density diagrams. Using several queueing models, speed is determined, based on different arrival and service processes. The exact shape of the different speed-flow-density diagrams is largely determined by the model parameters. Therefore we believe that a good choice of parameters can help to adequately describe reality. We illustrated this with an example, using the most general models (including a state dependent model) for a highway. REFERENCE Daganzo, C.F., 1997, Fundamentals of Transportation and Traffic Operations, Elsevier Science Ltd., Oxford. Heidemann, D., 1991, Queue length and waiting-time distributions at priority intersections, Transportation Research-B, 25, 163-174. Heidemann, D., 1994, Queue length and delay distributions at traffic signals, Transportation Research-B, 28B, 377-389. Heidemann, D., 1996, A queueing theory approach to speed-flow- density relationships, Transportation and Traffic Theory, Proceedings of the 13th International Symposium on Transportation and Traffic Theory, Lyon, France, 14-26 July 1996. Heidemann, D. and Wegmann H., 1997, Queueing at unsignalized intersections, Transportation Research - B, 31, 239-263.
Neues verkehrswissenschaftliches Journal - Ausgabe 16: Capacity Research in Urban Rail-Bound Transportation with Special Consideration of Mixed Traffic