Unit 13 - Queuing Analysis (Part 1)
Unit 13 - Queuing Analysis (Part 1)
INSTRUCTIONAL MATERIAL
IN HIGHWAY AND
RAILROAD ENGINEERING
(CEPC 313)
A.Y. 2020-2021
REFERENCES:
1. Garber, N., Hoel, L. 2009 Traffic and Highway Engineering Fourth Edition Toronto: CENGAGE
Learning
2. Mannering, F., Washburn, S., 2013 Principles of Highway Engineering and Traffic Analysis Fifth
Edition New York: John Wiley & Sons.
Unit 13: Queuing Analysis (Part 1)
OBJECTIVES
A queue is formed when arrivals wait for a service or an opportunity, such as the arrival of an accepted
gap in a main traffic stream, the collection of tolls at a tollbooth or of parking fees at a parking garage,
and so forth.
QUEUING NOTATION
The generally used notation for describing queue is given by X / Y / N where in place of X the arrival
distribution type should be used, in place of Y the service distribution type should be used, N
represents the number of servers.
Some of the most general notations are D/D/1, D/D/N, M/D/N, M/M/N, M/G/N where D stands for
deterministic, M stands Markovian, G stands for Generalized.
Traffic Intensity - Defining a new term (traffic intensity) for the ratio of average arrival to departure
rates as
(Eq. 13.1)
λ
ρ= where
μ
ρ = traffic intensity, unitless,
λ = average arrival rate in vehicles per unit time, and
μ = average departure or service rate in vehicles per unit time
OVERSATURATED AND UNDERSATURATED QUEUES
Oversaturated queues are those in which the arrival rate is greater than the service rate.
Undersaturated queues are those in which the arrival rate is less than the service rate.
The length of an undersaturated queue may vary but will reach a steady state with the arrival of units.
The length of an oversaturated queue, however, will never reach a steady state but will continue to
increase with the arrival of units.
A queuing model that assumes one departure channel and exponentially distributed departure times in
addition to exponentially distributed arrival times (an M/M/1 queue) is applicable in some traffic
applications.
Figure below is a schematic of a single-channel queue in which the rate of arrival is λ veh/h and the
service rate is μ veh/h. For an undersaturated queue, μ> λ assuming that both the rate of arrivals and
the rate of service are random, the following relationships can be developed:
Probability of n units in the system, P(n):
(Eq. 13.2)
P(n)= ρn (1−ρ)
The expected number of units in the system, E(n):
(Eq. 13.3)
λ
E(n)=
μ−λ
The expected number of units waiting to be served (that is, the mean queue length) in the system,
E(m):
(Eq. 13.4)
λ2
E( m)=
μ( μ−λ)
Note that E(m) is not exactly equal to E(n) – 1, the reason being that there is a definite probability of
zero units being in the system, P(0).
ρ2
σ=
1−ρ
ILLUSTRATIVE PROBLEMS
PROBLEM 1
On a given day, 425 veh/h arrive at a tollbooth located at the end of an off-ramp of a rural expressway. If
the vehicles can be serviced by only a single channel at the service rate of 625 veh/h, determine
PROBLEM 2
An off-ramp traffic must stop at a single tollbooth at the entrance of the Highway 78. The arrival rate at
the tollbooth is 80 vehicles per hour, if the service rate is 86 vehicles per hours, determine the number
of vehicles waiting to be served.
PROBLEM 3
Vehicles arrive at the entrance booth of MGM Hotel in Las Vegas to see the boxing match of Manny
Pacquiao at the rate of 4 vehicles per minute. It took the attendant 30 seconds to process the parking
tickets. Determine the rate in which the vehicles move after getting their tickets.
PROBLEM 4
Vehicles arrived at the entrance of San Juanico bridge in Tacloban at a rate of 180 vehicles per hour.
Vehicles must stop at the toll gate where vehicles have to pay a toll fee upon entering the bridge. It
would take 15 sec. per vehicle for the attendant to issue a trip ticket. Compute the traffic intensity at the
bridge.
PROBLEM 5
Vehicles arrived at a stop sign at the corner of P. Del Rosario St. and Roxas Blvd. at an average rate of
250 vehicles per hour. Average waiting time at the stop sign is 12 sec per vehicle. If both arrivals an
departure are exponentially distributed, what is the average length of queue?
PROBLEM 6
On a given day, 425 vehicles per hour arrive at a toll booth located at the end of an off-ramp of the
South Super Highway. If the vehicles can be serviced by only a single channel at the service rate of 625
vehicles per hour, determine the average number of vehicles in the system.
PROBLEM 7
The number of vehicles that can enter the on-ramp of Subic-Clark Expressway is controlled by a
metering system. If the vehicles can enter the expressway at a rate of 500 vehicles per hour, and the
rate of arrival of the vehicles at the on-ramp is 400 vehicles per hour during the peak hour. Determine
the expected number of units waiting to be served in the system.
PROBLEM 8
Off-ramp traffic must stop at a single tollbooth. The arrival rate of the tollbooth is 45 vehicles per hour. If
the service rate is 60 vehicles per hour, determine the number of vehicles waiting excluding the vehicles
being served.
PROBLEM 9
Passenger cars arrive at the stop sign of an average rate of 280 per hour. Average waiting time at the top
sign is 12 sec. If both arrivals and departure are exponentially distributed, what would be the average
delay per vehicle in minutes. Assume both arrival and departure rates are exponentially distributed.
PROBLEM 10
Cars have a long queue on a stop sign. If the total time delay of queuing is one minute, compute the
traffic density if the service rate of the road is only 360 vehicles per hour. Assume both arrival and
departure rates are exponentially distributed.