0% found this document useful (0 votes)
13 views24 pages

Traffic Assignment in Transport Systems

The document discusses transport network modeling and traffic assignment, focusing on real-world implications such as infrastructure development and traffic flow changes. It covers concepts like User Equilibrium (UE), System Optimum (SO), and the Braess Paradox, along with mathematical relationships governing traffic flow and travel time. Various examples illustrate how to determine optimal routes and tolls to improve traffic conditions and revenue.

Uploaded by

alanzhi1
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
13 views24 pages

Traffic Assignment in Transport Systems

The document discusses transport network modeling and traffic assignment, focusing on real-world implications such as infrastructure development and traffic flow changes. It covers concepts like User Equilibrium (UE), System Optimum (SO), and the Braess Paradox, along with mathematical relationships governing traffic flow and travel time. Various examples illustrate how to determine optimal routes and tolls to improve traffic conditions and revenue.

Uploaded by

alanzhi1
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

CIVL 2700/9700

Transport Systems
Topic 2: Transport Network Modelling (Traffic Assignment)

Reza Mohajerpoor

CIVL2700: Transport Systems - Week 2 1


Relevance to Real-world problems
Ø What are the transport implications of building a new infrastructure (e.g. WestConnex)?

Ø How the traffic situation (flow) will change if M4 is no longer tolled?

Ø Which street is the best to install an advertisement billboard on to maximize the user
view?

Ø Which bus route (e.g. bus number 438 perhaps) is the best to consider for
advertisement-on-bus to maximize the user view?
CIVL2700: Transport Systems - Week 2 2
Relevance to Real-world problems
Ø Assuming a higher usage of Electric Vehicles (EVs) in Sydney, where should the charging
stations be built?

Ø Which street to be tolled in CBD to maximize the toll revenue of the City of Sydney?

Ø Which street to be tolled in CBD to minimize the total delay time in CBD?

CIVL2700: Transport Systems - Week 2 3


Agenda
• Introduction
• Performance Function
• User Equilibrium (UE)
• Braess Paradox
• System Optimum (SO)
• Congestion Pricing

CIVL2700: Transport Systems - Week 2 4


Introduction
• Human’s decision making in transport systems

CIVL2700: Transport Systems - Week 2 5


Introduction
• Let us assume total traffic demand [veh/time] between Origins (O)
and Destinations (D) are known x1

x2
• Traffic assignment (route choice) question is:
Which routes are utilized to what extent [veh/time]?

• Decisions are primarily a function of route travel time that are


determined by route traffic flow

• The mathematical relationship between route travel time and route


traffic flow is called a Performance Function

• Classic equilibrium problem (traffic assignment)


CIVL2700: Transport Systems - Week 2 7
Performance Function
• The mathematical relationship between route travel time
and route traffic flow is called a Performance Function

• The simplest relationship: linear


• More realistic: nonlinear relationship

CIVL2700: Transport Systems - Week 2 8


Theory of User Equilibrium
• Two important assumptions
i. travellers select routes based only on route travel times
ii. travellers know all the travel times on all the available routes

• User equilibrium (UE):


o when individual travellers cannot improve their travel times by unilaterally
changing routes
o “The travel time between a specified origin and destination on all used routes is
equal, and less than or equal to the travel time that would be experienced by a
traveller on any unused route” (Wardrop)

CIVL2700: Transport Systems - Week 2 9


Example
Two highways serve a busy corridor with a traffic demand that is fixed at 6000 vehicles
during the peak hour (1 hour). The performance functions for the two routes are

t1 = 4 + 5(x1/c1) and t2 = 3 + 7(x2/c2)

where t’s are in minutes and flows (x’s) and capacities (c’s) are in vehicles per hour. Initially,
the capacities of routes 1 and 2 are 4400 veh/h and 5200 veh/h, respectively.

a. Find the total travel time on the network assuming that UE condition hold.
b. If a highway reconstruction project cuts the capacity of route 2 to 2200 veh/h, how many
additional vehicle hours of travel time will be added in the corridor assuming that UE
condition hold?
CIVL2700: Transport Systems - Week 2 10
Example (solution 1)
a. Find the total travel time on the network assuming that UE conditions hold.
At UE and if both routes are used then 𝑡! = 𝑡" The number The number
of vehicles of vehicles
𝑥!+ 𝑥" = 6000 [veh/h] using route using route
𝑥! 𝑥" 2 during the 1 during the
4+5 =3+7 1 peak hour 1 peak hour
𝑐! 𝑐"
6000 − 𝑥" 𝑥" Total travel time is:
4+5 =3+7 𝑡! 2851 ×2851×1
4400 5200
30000 5𝑥" 7𝑥" + 𝑡" 3149 ×3149×1
1+ − = = 7.24×2851 + 7.24×3149
4400 4400 5200
7.82 = 0.00248𝑥" = 43440 [[Link]]
𝑥" ≈ 3149[veh/h]
𝑥! ≈ 2851 [veh/h] both routes are used
CIVL2700: Transport Systems - Week 2 11
Example (solution 2)
b. If a highway reconstruction project cuts the capacity of route 2 to 2200 veh/h, how many
additional vehicle hours of travel time will be added in the corridor assuming that UE
condition hold? The number of The number of
vehicles using route 2 vehicles using
At UE and if both routes are used then 𝑡! = 𝑡" during the 1 peak hour route 1 during
𝑥! + 𝑥" = 6000 Total travel time is: the 1 peak hour
𝑥! 𝑥"
4+5 =3+7 𝑡! 4189 ×4189×1
𝑐! 𝑐"
6000 − 𝑥" 𝑥" + 𝑡" 1811 ×1811×1
4+5 =3+7
4400 2200 = 8.76×4189 + 8.76×1811
30000 5𝑥" 7𝑥" = 52560 [[Link]]
1+ − =
4400 4400 2200
7.82 = 0.00432𝑥" Additional hours is:
𝑥" ≈ 1811 [veh/h] 52560 − 43440 = 9120 [[Link]]
𝑥! ≈ 4189 [veh/h] both routes are used #!"$
= 152 [veh.h]
%$
CIVL2700: Transport Systems - Week 2 12
Example
Two routes connect a city and a suburb. During the peak-hour (1 hour) morning
commute, a total of 4500 vehicles travel from the suburb to the city. Route 1 has a 100
km/h speed limit and is 10 km in length; route 2 is 5 km in length with a 75 km/h speed
limit. Studies show that the travel time on route 1 increases two minutes for every
additional 500 vehicles added. Minutes of travel time on route 2 increase with the
square of the number of vehicles, expressed in thousands of vehicles. Determine user-
equilibrium travel times.

CIVL2700: Transport Systems - Week 2 13


Example (solution)
Free flow Route 1: 10/100*60=6 [min]
Free flow Route 2: 5/75*60=4 [min]
𝑡! = 6 + 4𝑥!
𝑡" = 4 + 𝑥""
𝑞 = 𝑥! + 𝑥" = 4.5
First: to determine whether or not both routes are used
6 + 4𝑥! = 4 + 𝑥""
6 + 4(4.5 − 𝑥") = 4 + 𝑥"" Ø What is the value of q’?
𝑥" = 2.899 [103 veh] Ø If less than q’ vehicles travel
𝑥! = 1.601 [103 veh] during the 1 hour peak period,
what are the user equilibrium
𝑡! = 𝑡" = 12.4 [min] travel times?
Total travel time during the peak period (UE):
4500×12.4 = 55800 [[Link]]
CIVL2700: Transport Systems - Week 2 14
Example: the Braess Paradox (1)
Given the original network, determine UE travel times.

CIVL2700: Transport Systems - Week 2 15


Example: the Braess Paradox (2)
Now, city council has built a new road (even an
expensive one, e.g. a tunnel).
Given the new network, determine UE travel times.

Implication?

CIVL2700: Transport Systems - Week 2 16


Theory of System Optimum
• Two important assumptions
i. travellers select routes based only on route travel times
ii. travellers know all the travel times on all the available routes

• System Optimum (SO):


o the total travel time for the overall network is minimum.
𝑡! : link n travel time
min 𝑆 𝑥 = ; 𝑥0 𝑡0 𝑥! : the number of vehicle travelling on link n
/
0
subject to
𝑥0 ≥ 0
Inflow=outflow at each node
CIVL2700: Transport Systems - Week 2 17
Example
Two routes connect a city and a suburb. During the peak-hour (1 hour) morning
commute, a total of 4500 vehicles travel from the suburb to the city. Route 1 has a 100
km/h speed limit and is 10 km in length; route 2 is 5 km in length with a 75 km/h speed
limit. Studies show that the travel time on route 1 increases two minutes for every
additional 500 vehicles added. Minutes of travel time on route 2 increase with the
square of the number of vehicles, expressed in thousands of vehicles. Determine the
system-optimum travel time.

CIVL2700: Transport Systems - Week 2 18


Example (solution)
min 𝑆 𝑥 = 𝑥!𝑡! 𝑥! + 𝑥"𝑡" 𝑥"
/! ,/"
s.t.
𝑞 = 𝑥! + 𝑥" ; 𝑥!, 𝑥" ≥ 0

𝑆 𝑥 = 6𝑥! + 4𝑥!" + 4𝑥" + 𝑥"5 = 𝑥"5 + 4𝑥"" − 38𝑥" + 108


d𝑆(𝑥)
= 3𝑥"" + 8𝑥" − 38 = 0
d𝑥"
𝑥" = 2.467 [103 veh]
𝑥! = 2.033 [103 veh] Total travel time during the peak period (SO):
𝑡! = 14.13 [min] 2033×14.13 + 2467×10.09 = 28726.3 + 24892.0 = 53618.3 [[Link]]
Compare this with UE total travel time of 55800!
𝑡" = 10.08 [min]

CIVL2700: Transport Systems - Week 2 19


Congestion Pricing
• User equilibrium (UE):
o when individual travellers cannot improve their travel times generalized travel cost
by unilaterally changing routes
o The travel time generalized travel cost between a specified origin and destination on
all used routes is equal, and less than or equal to the travel time generalized travel
cost that would be experienced by a traveller on any unused route

CIVL2700: Transport Systems - Week 2 20


Example
Two routes connect a city and a suburb. During the peak-hour (1 hour) morning
commute, a total of 4500 vehicles travel from the suburb to the city. Route 1 has a 100
km/h speed limit and is 10 km in length; route 2 is 5 km in length with a 75 km/h speed
limit. Studies show that the travel time on route 1 increases two minutes for every
additional 500 vehicles added. Minutes of travel time on route 2 increase with the
square of the number of vehicles, expressed in thousands of vehicles. If the value of time
is $1 per minute and a toll of $1 is charged for each trip on route 2 determine the travel
times of each route assuming drivers’ behaviour follow generalized travel cost UE
condition.

CIVL2700: Transport Systems - Week 2 21


Example (solution)
𝑡! = 6 + 4𝑥! How much money the city
𝑡" = 4 + 𝑥"" makes in this example?
𝑞 = 𝑥! + 𝑥" = 4.5
First: to determine whether or not both routes are used What is the optimal toll value?
1× 6 + 4𝑥! = 1× 4 + 𝑥"" + 1
6 + 4(4.5 − 𝑥") = 5 + 𝑥"" Repeat the problem with $1
toll for route 1.
𝑥" = 2.796 [103 veh]
𝑥! = 1.704 [103 veh]
𝑡! = 12.8 [min]
𝑡" = 11.8 [min]
Total travel time during the peak period (generalized travel cost UE):
1704×12.8 + 2796×11.8 = 54804 [[Link]]
CIVL2700: Transport Systems - Week 2 22
UE, SO, Toll Examples (solution)

UE SO Toll ($1) Optimal toll ($ ??)


𝑥2 1601 2033 1704 2033
𝑥3 2899 2467 2796 2467
𝑡2 12.4 14.13 12.8 14.13
𝑡3 12.4 10.08 11.8 10.08
TTT 55800 53618 54804 53618

CIVL2700: Transport Systems - Week 2 23


Thoughts
• Large-scale networks (multiple O-D, extensive route options)
• Multimodal flows (e.g. trucks, buses)
• Stochastic TA: Different cost criteria; perception errors, limited information;
habits
• Static vs. Dynamic Equilibrium
o Time varying demands
o Changing traffic conditions in the network
o Real-time drivers’ diversion (e.g. Incidents)

CIVL2700: Transport Systems - Week 2 24

You might also like