You are on page 1of 16

CIV 475 Traffic Engineering

Queuing Theory

Mannering / Kilareski
Chapter 5

OREGON TECH CIV475


Lindgren 1
Queue

A ‘queue’ is simply a line


There were 16 cars in line at the toll
booth
The toll booth queue was 16 cars

OREGON TECH CIV475


Lindgren 2
Queuing Theory

Queuing theory is a broad field of study of


situations that involve lines or queues
retail stores
manufacturing plants
transportation
traffic lights
toll booths
stop signs
etc.
OREGON TECH CIV475
Lindgren 3
Queuing Theory -
acronyms
FIFO - a family of models that us the
principle of “first in first out”
LIFO - “last in first out”
a/d/N notation (aka Kendall notation)
a - arrival type (either D- deterministic, or M-
mechanistic)
d - departure type (either D- deterministic, or M-
mechanistic)
N - number of “channels”
OREGON TECH CIV475
Lindgren 4
D/D/1 FIFO
 Entrance gate to National Park
 Deterministic arrivals and departures, one fee booth, first in
first out
 At the opening of the booth (8:00am), there is no queue,
cars arrive at a rate of 480veh/hr for 20 minutes and then
changes to 120veh/hr
 The fee booth attendant spends 15seconds with each car
 What is the longest queue? When does it occur?
 When will the queue dissipate?
 What is the total time of delay by all vehicles?
 What is the average delay, longest delay?
 What delay is experienced by the 200th car to arrive?
OREGON TECH CIV475
Lindgren 5
More than D/D/1

While D/D/1 queuing is easy to understand and


graphical solutions are available, it may not be
the best model to use in traffic situations since
arrivals are not Deterministic (as you will see by
collecting data on some real traffic streams)
Derivation of Stochastic (Mechanistic) queuing
equations is beyond the scope of this course,
but the equations are listed in the text book
read up on M/D/1, M/M/1 and M/M/N
OREGON TECH CIV475
Lindgren 6
Queuing at traffic lights

OREGON TECH CIV475


Lindgren 7
Graph of Flow vs. time
Red is shown as darker gray, green is lighter gray

Constant Arrival Flow

OREGON TECH CIV475


Lindgren 8
During the red interval for the approach,
vehicles cannot depart from the intersection and
consequently, a queue of vehicles is formed.
When the signal changes to green, the vehicles
depart at the saturation flow rate until the
standing queue is cleared.
Once the queue is cleared, the departure flow
rate is equal to the arrival flow rate.
Departure flow rates are shown in the next
figure
OREGON TECH CIV475
Lindgren 9
Departure Flow Diagram

OREGON TECH CIV475


Lindgren 10
Sketch a graph showing how the queue
length changes with time during a red-
green period for one movement of an
intersection

OREGON TECH CIV475


Lindgren 11
During the red interval, the queue of vehicles
waiting at the intersection begins to increase.
The queue reaches its maximum length at the
end of the red interval
When the signal changes to green, the queue
begins to clear as vehicles depart from the
intersection at the saturation flow rate

OREGON TECH CIV475


Lindgren 12
There is another graph that allows us to glean
even more information from our model.
Imagine a plot where the x-axis is time and the
y-axis contains the vehicle numbers according to
the order of their arrival.
Vehicle one would be the first vehicle to arrive
during the red interval and would be the lowest
vehicle on the y-axis.
If you were to plot the arrival and departure
(service) times for each vehicle, you would
get a triangle OREGON TECH CIV475
Lindgren 13
OREGON TECH CIV475
Lindgren 14
For a given time, the difference between the
arrival pattern and the service pattern is the
queue length.
For a given vehicle, the difference between the
service pattern and the arrival pattern is the
vehicle delay.
In addition, the area of the triangle is equivalent
to the total delay for all of the vehicles.

OREGON TECH CIV475


Lindgren 15
Assignment

Text problems: 5.4, 5.9, 5.11, 5.14, 5.22


 Collect traffic arrival data (30-1minute increments) on a
moderately congested road that is away from the
influence of traffic signals, plot a histogram (figure 5.5)
of your data, on the same histogram show a Poisson
distribution model of the same # of arrivals, determine if
your data follows a Poisson distribution
 Collect 30 minutes of headway data, plot as per (figure
5.6), on the same graph, plot the exponential (Poisson)
probabilities, determine if your data follows a Poisson
distribution
OREGON TECH CIV475
Lindgren 16

You might also like