Transportation Modeling

Dr. Md. Jahir Bin Alam

 Transportation Planning

 Transportation planning is a preparation planning to move

/transfer human, animal or goods to one place to another
place
 The planning will related to the operation of the highway
system, geometry and operation of traffic facilities.

Kadiyali, L. R. ―Traffic Engineering and Transport Planning‖, Khanna Publishers

Function of Transportation
Planning
 To identify highway system components
 To define transportation planning
 To recall the process and purpose of construction
planning design
 To identify models of transportation such as gravity
model, logit model etc.
Characteristic of
Transportation Planning
 Determine the transportation needs
 Built transportation formulas
 Study the profitability
 Travel pattern should be clear, stable and can be
controlled
 Relationship between different types of modes of
transport
 Transportation system can influence on
development of the area and ready to serve the
area
Important of transportation
Planning
 Efficiency---To achieve efficient
management and better management of
existing resources
 Quality---to reduce a negative impact to
traffic which produce pollution
 Equity--- to meet travel demand
 Definition

Transportation modeling is a science that seeks to study the problems that

arise in providing transportation facilities in an urban, regional, national
setting and to prepare a systematic basis for planning such facilities.
Four Steps of Conventional
Transportation Modeling
Trip Generation
-- decision to travel for a specific purpose (eat
lunch)
Trip Distribution
-- choice of destination (a particular restaurant?
The nearest restaurant?)
Mode Choice
-- choice of travel mode (by bike)
Network Assignment
-- choice of route or path
Travel behavior modeling/Demand modeling
Travel behavior modeling

Three types
 Activity based model
 Tour based model
 Trip based model
Steps of Transportation Demand Modeling

There are four steps-

 Trip Generation (How much people goes from a specific area/location?)

 Trip Distribution (How much people distributed per location?)
 Modal split (Which vehicle will be used?)
 Trip Assignment (Which road will be used?)
 What is trip generation?

It is the process by
which measures of urban
activity are converted into
numbers of trips.
In trip generation, the
planner attempts to
quantify the relationship
between urban activity and
travel.
Three ways for estimating the number of trips

Regression models Y = dependent var.

(trips/household)
X1, X2, etc. =
independent variables
Trip rates, like # of trips/1000ft2
Category analysis (cross-classification analysis)
• Trip production:
– p = trip purpose
– i = zone
– h = household type grouping
– ai(h) = number of households of type h
in zone i
– tp(h) = trip rate for trip of type p for
households of type h
 1. Trip Generation:

Def: Total numbers of trips that produced from a study area and are attracted
by a study area.

Factors affecting trip production:

 No. of population.
 Ownership of car.
 Family size and composition.
 Income.
 Land use characteristics.

Factors affecting trip attraction:

 Retail floor space.
 Employment opportunity.
 Land use characteristics.
Attraction
Number and types of retail
facilities
Number of employees
Land use
Production
Car ownership
Income
Population (employment
characteristics)
 1. Trip Generation (Cont.)

Methodology

Following methods is used for trip generation modeling

Cross-classification
 Regress analysis. Covered in 355
Multiple regression analysis
 Growth factor. Mathematical equation that describes trips as a function of another variable
Similar in theory to trip rate
Won’t go into
 Cross classification. Trip-rate analysis models
Average trip-production or trip-attraction rates for specific types of producers and attractors
More suited to trip attractions

1. Regression analysis:
Pi=C+B1x1+B2x2+…………………+Bnxn
Ai= C‌'‌ +B’1x’1+B’2x’2+…………………+B’nx’n
where, Pi= trip production
Ai = trip attraction
C1, B1, B2………………Bn = co-efficient of trip production
x1, x2………….xn = trip production independent variable.
C’1, B’1, B’2………………B’n = co-efficient of trip attraction
x’1, x’2………….x’n = trip attraction independent variable.
1. Trip Generation(Cont.)
Example of trip Generation Model

Trip production, PM = 0.62 POP + 0.12 AMO

Trip attraction AM = 0.41 RFS + 0.08 EMO

Where, POP = population; AMO = automobile ownership;

RFS = retail floor space; EMO = employment opportunity.
For 100 employees in a retail shopping center, calculate
the total number of trips
Home-based work (HBW) =
100 employees x 1.7 trips/employee = 170
Home-based Other (HBO) =
100 employees x 10 trips/employee = 1,000
Non-home-based (NHB) =
100 employees x 5 trips/employee = 500

Total = 170 + 1000 + 500 = 1,670 daily trips

force total trip attractions to equal total trip productions
Pi = trips produced by zone i
Ai = total trips attracted by zone i

• The adjusting factor to adjust the attractions

2. Trip Distribution

It determines where the trips produced in each zone will go, how they will
be divided among all other zones.
2. Trip Distribution (Cont.)

Factors affecting
 Trip production and attraction
 Travel time
 Travel cost
 Distance

Methodology
 Gravity Model
 Growth factor Model
 Intervening Opportunity Model
The gravity model assumes that the trips produced at an
orgin and attracted to a destination are proportional to
The total trips productions at the orgin and the total attraction
At the destination. Frictional factors represents the
impedence of persons to make trips of the various
durations or distances. It indicates travel time increase.
2. Trip Distribution (Cont.)
 2. Trip Distribution (Cont.)

Problem of Gravity Model:

Calculate trip distribution matrix using gravity model. Necessary information
are given below
Travel time (min)
Zone 1 2 3
1 0 5 7
2 4 0 6
3 6 6 0

Zone 1 2 3 total
Trip production 140 330 280 750
Trip attraction 300 270 180 750
2. Trip Distribution (Cont.)

Solution: Zone 1 Zone 2

Zone 3

Trip Distribution:
T1-2=140[(270/5²)/(270/52+180/7²)]=104.46≈104.

T1-3=140[(180/7²)/(270/5²+180/7²)]=35.53≈36.

T2-1=330[(300/4²)/(300/4²+180/6²)]=260.52≈261.

T2-3=330[(180/6²)/(300/4²+180/6²)]=69.47≈69.

T3-1=280[(300/6²)/(300/6²+270/6²)]=147.36≈147.

T3-2=280[(270/6²)/(300/6²+270/6²)]=132.63≈133.
2. Trip Distribution (Cont.)

Initial trip distribution matrix

Zone 1 2 3 Total production

1 0 104 36 140
2 261 0 69 330
3 147 133 0 280
Attraction 408 237 105
Actual attraction 300 270 180

Dissimilarity appeared in trial trip attraction with actual trip attraction.

So further trial in needed
2. Trip Distribution (Cont.)

Trial 1
Zone 1 2 3 Production Actual
production
1 0 270/237*104 180/105*36 180 140
=118 =62
2 300/408*261 0 180/105*69= 310 330
=192 118
3 300/408*147 270/237*133=1 0 260 280
=108 52
Attraction 300 270 180

Dissimilarity appeared in trial trip production with actual trip production.

So further trial in needed
2. Trip Distribution (Cont.)

Trial 2

Zone 1 2 3 Production

1 0 140/180*118=92 140/180*62=48 140

2 330/310*192=204 0 330/180*118=126 330

3 280/160*108=116 280/260*152 0 280

=164
Attraction 320 256 174

Trial trip attraction is very close to actual attraction. So further trial is not
required.
Assignment
K=1
3. Mode-choice: Modal Split Model

These are primarily oriented towards predicting the percentage of

individuals who will choose one mode over other for making a particular
trip.

Factors influencing trip choice are

 Trip Characteristics (trip purpose, trip length, time of the day etc.)

 Trip maker characteristics (family income, age, auto ownership etc.)

 Transportation system characteristics (Travel time, travel cost etc.)

Factors affecting mode
choice
 Factors that may explain a trip maker’s choosing a specific
mode of transportation for a trip are grouped----
(A) Trip maker’s characteristics: Income, car ownership, car
availability, age
(B) Trip characteristics---- trip purpose (work, shopping,
recreation, destination orientation, trip length)
(C) Transportation system characteristics—waiting time,
speed, cost, comfort, convenience, access to terminal or
transfer location.
Methodology

 Discrete choice model

 Logit Model
Binary logit: Two alternatives i.e car and bus

Multinomial logit: More than two alternatives. i.e bus, car, auto
rickshaw

Nested logit: If alternatives are correlated. i.e Motor vehicle, Non

motor vehicle

Cross nested logit: Each alternative may belongs two more than one
nest

 Probit Model
Probit model is based on normal distribution only. It considers
cumulative normal distribution. It does
not have a closed form. The choice of probability is an integral
Logistic distribution is used because it approximates normal
Distribution quite well. Gumma distribution can als be justified.
It is analytically convenient to use logit model. Logit model
have faller tails than a normal distribution.

Logit =
log { Normal
y/(1- distribuion
y)}
The logit model trades off the relative utilities
of various modes.

Logit analysis assumes that probability of the

occurrence of an event varies w.r. to function
as a sigmoid curve (logistic curve).
Probit analysis is based on the principle that if members of
population are subjected to a stimulus that can range over an
infinite scale, the frequency of response to stimulus will be
normally distribution.
Logit analysis assumes that probability of the occurrence of
an event varies with respect to function as sigmoid curve
(logistic curve).
Modal Split Model (Cont.)

Methodology
Logit Model:

Pit =

Where
Pit = Probability of individual t choosing mode i.
Uit = Utility of mode i (car) for a individual t.
Ujt = Utility of mode j (bus) on individual t.

Ucar = β1 (travel time) + β2 (travel cost)

Ubus = β1 (travel time) + β2 (travel cost)
β1 and β2 can be evaluated by BIOGEM software.
Consumer maximum utility
U(bus) > U (auto)-----choice bus

P (y=1) = 1/(1+e –(β0 + β1x) )

Modal Split Model (Cont.)

Model Interpretation
 Problem of Logit Model:

Utility function of road user travelling from sub urban to CBD of Network City
using mode auto and bus are given below:
U auto= -0.06-0.07(auto travel time in min.)-0.105(parking cost in dollar) +
0.00001(HH income in dollar)
U bus= -0.07(bus travel time in min)-0.12(bus waiting time in min.)-0.66(bus fee
in dollar)
Variable Auto Bus
Auto travel time in min. 40 -
Parking cost (dollar) 8 -
Bus fare (dollar) - 2
Bus travel time (min) - 60
Bus head way (min) - 20
Annual HH income $ 8000
1. Compute the probability as individual will choose auto for trip?
2. In order to increase rider ship in public transport which option will
consider transport authority?
 Reducing bus fare 50 cents
 Doubling the no of bus
 Doubling parking price
Which option is more effective to increase public transport ridership?
3. Value of time of bus ?
 Utility function: measures satisfaction derived from
choices
 Disutility function: represents generalized costs of
each choice
 Usually expressed as the linear weighted sum of the
independent variables of their transformation
U = a0 + a1X1 + a2X2 + ….. + arXr

U: utility derived from choice

Xr: attributes
ar: model parameters
 Calculates the probability of selecting
a particular mode

p(K) = ____eUk__
 eUk

p: probability of selecting mode k

 Example 1:
Travel characteristics between two zones

Utility functions for auto and transit

U = ak– 0.35t1 – 0.08t2 – 0.005c
ak = mode specific variable
t1 = total travel time (minutes)
t2 = waiting time (minutes) Variable Auto Transit
ak -0.46 -0.07
c = cost (cents)
t1 travel 20 30
waiting
t2 8 6
c 320 100
Find the utility value for different mode and comment
on choice of mode.
Find out probability of choice of different mode
Uauto = -0.46 – 0.35(20) – 0.08(8) – 0.005(320) = -9.70

Utransit = -0.07 – 0.35(30) – 0.08(6) – 0.005(100) = -11.55

Uauto = -9.70

Utransit = -11.55

Logit Model:

p(auto) = ___eUa __ = _____e-9.70 ____ = 0.86

eUa + eUt e-9.70 + e-11.55

p(transit) = ___eUt __ = _____e-11.55 ____ = 0.14

eUa + eUt e-9.70 + e-11.55
 Example 2:

The city decides to spend money to create and improve bike trails so
that biking becomes a viable option, what percent of the trips will be
by bike?
Assume:
• A bike trip is similar to a transit trip
• A bike trip takes 5 minutes more than a transit trip but with no
waiting time
• After the initial purchase of the bike, the trip is ―free‖
Travel characteristics between two zones

Variable Auto Transit Bike

ak -0.46 -0.07 -0.07
Find out probability of choice of
t1 Travel time 20 30 35 different modes.

t2 Waiting time 8 6 0
c 320 100 0
 Uauto = -0.46 – 0.35(20) – 0.08(8) – 0.005(320) = -9.70

Utransit = -0.07 – 0.35(30) – 0.08(6) – 0.005(100) = -11.55

Ubike = -0.07 – 0.35(35) – 0.08(0) – 0.005(0) = -12.32

Uauto = -9.70, Utransit = -11.55, Ubike = -12.32

Logit Model:

p(auto) = _____eUa ____ = _______e-9.70 ______ = 0.81

eUa + eUt +eUb e-9.70 + e-11.55 + e-12.32

p(transit) = _____eUt__ __ = ______e-11.55 ______ = 0.13

eUa + eUt +eUb e-9.70 + e-11.55 + e-12.32

p(bike) = _____eUt__ __ = ________e-11.55 ______ = 0.06

eUa + eUt +eUb e-9.70 + e-11.55 + e-12.32
4 Trip Assignment

Objectives:
 In which route people will travel.

Parameter/Input:
 Travel time, travel cost, numbers of trip between zone pairs etc.

Method:
 All or nothing assignment model
 Capacity restraint assignment.
All-or-nothing Assignment
This is the simplest technique and is based on the premise that the
route followed by traffic is one having the least travel resistance.
The resistance itself can be measured in term of travel time,
distance, cost, safety, reliability, convenience.

Capacity Resistant Assignment: It is a process in which the travel resistance of a

link is increased according to a relationship between practical capacity of link and
volume assigned to the link.

Considering capacity and

congestion effects
Problem of Trip Assignment:

An urban area is divided into 4 zones for transportation demand analysis.

Zone pair trip distribution is given in table and zone pair travel time is given
in fig bellow. Calculate total no. of trip assigned in each link of road. Using all
or nothing assignment model.

OD 1 2 3 4
1 0 100 200 300
2 200 0 200 400
3 100 300 0 100
4 200 400 200 0
 Solution:

OD Minimum 1-2 1-3 1-4 2-4 3-4

path travel
1-2 1-2 100
2-1 2-1 200
1-3 1-3 200
3-1 3-1 100
1-4 1-4 300
4-1 4-1 200
2-3 2-1-3 200 200
3-2 3-1-2 300 300
4-3 4-3 200
considered 3-4 3-4 100
4-2 4-1-2 400 400
2-4 2-1-4 400 400
Total 1600 800 1300 300
Sustainable transportation
Sustainable transportation should promote sustainable
development.
Sustainable development means development that meets the needs
of the present without compromising the ability of future generation
to meet their own needs.

Sustainability: A set of environmental, economic and social

conditions in which all of the society has the capacity and the
opportunity to maintain and improve its quality of life indefinitely
without degrading the quality, or availability of natural, economic
and social resources.
 Environmental performance
 Economic development
 Social inclusion
Haddon Matrix
Element Before crash In crash After crash

Human Training/Education, Behavior Personal protection Emergency

(drink equipment medical
driving), Attitudes, Licensing, Helmet, Seatbelt etc service
Motorist physical impairment,
Rider risk recognition and
judgment, Crash
avoidance skills

Vehicle Horns, Braking, ITS, Vehicle Secondary safety to Salvage

Design, Passenger/Load, impact
Vehicle Equipment, Vehicle protection
Safety Equipment, Vehicle
Energy absorbing
Performance
steering columns
Head restraints
Forgiving interior
fittings

Road Road geometry Road side safety Restoration of

Visibility, Road Barriers road and
surface condition, traffic
Road safety audit device
Drainage system in Highway:
Principles of good drainage:
1. The surface runoff over the pavement surface and the
shoulder should be drain as quickly as possible.
2. Precipitation over the open land adjoining the highway
should be led away from the highway through natural channel
or cross channel.
3. Landslide prone zone drainage system across embankment
water edged and flood prove zone need special investigation
for improving drainage.
4. Seepage and sub-surface water is deter metal to stability of
cut slopes effective system of sub surface drainage in required.
Rainfall duration is chosen
based on time of concentration
Q 1. A small catchment area consists of 2 km 2 of forest area (C=0.1);
1.20km 2 of cultivated area (C=0.2) and 1km 2 under grass cover (C=0.35) a
water course falls by 20m in a length of 2km.

IDF ---I = 80* T 0.2/(t +12) 0.5 where I = cm/hr, t = min T in year
Estimate peak rate of runoff a 25 year.

Solution:
C e = Sum of C I A I / (A I + A2 +A3)= (2 * 0.1 +1.2 *0.2
+1+0.35)/(2+1.2+1) = 0.1880952

t = 0.01947 (sqrt L 3/OH) 0.77 = 0.01947 (sqrt (2000 3/20) 0.77 =39.91min

IDF --- i = 80 T 0.2 / (39.91 +12) 0.5 = 21.13 cm/hr = 211.3 mm/hr
Q = C * i* A/3.6 = 0.18809* 211.3 *4.2 * 10 6 m 3/sec = 12.88 m 3/sec

d/D = 0.5(1-cos θ/2) where d is depth of water in drain

Thanks
To
All