01 general introduction of travel demand analysis dr binh

Contents of today’s class Introduction  Process of travel demand forecast  Travel demand models...  Travel demand analysis is not transportation planning; it can only support plann

General introduction of travel demand analysis

October 5, 2016 Transportation Planning and Policy

Contents of today’s class

 Introduction

 Process of travel demand forecast

 Travel demand models

Introduction

Characteristics of travel

 Each consumer or company makes a complex set of decisions relating the movement based on his/her

needs and environment

 These include purpose, frequency, timing, destination, and mode of trips

 Further, these decisions must be analyzed in the

context of the intertemporal behavior of the

consumer and company, and long-run decisions on

home, workplace, office location and on automobile ownership

 Thus, travel is a concomitant of various consumption and supply activities

What is the travel demand?

movements of individuals and/or goods.

example,

 Total number of passengers during a specific

period (ex Passenger per day)

 Total travel distance (ex Passenger kilometer)

 The number of goods transported by a specific

transportation mode

 The number of passengers traveling from one

place to other place

What is the travel demand

analysis?

 Travel demand analysis examines the characteristics of travel demand.

 Travel demand analysis is not

transportation planning; it can only

support planning, and in a few cases it may have the most important role in

the process.

Travel demand forecast

 Travel demand analysis can contribute to the travel demand forecast

Travel demand forecast in a

transportation planning

Travel demand forecast

Philosophical background of

travel demand analysis

 Empiricism is a theory of knowledge which emphasizes

those aspects of scientific knowledge that are closely related to evidence, especially as formed through deliberate experimental arrangements

 It is a fundamental requirement of scientific method that

all hypotheses and theories must be tested against observations of the natural world, rather than resting solely on a priori reasoning, intuition, or revelation.

 Scientific laws describe the general patterns of our

experiences To explain the phenomena in a scientific way means to show an example of scientific laws

 We can forecast the future of the same phenomena by

knowing the laws.

Travel demand analysis as a

part of social science

 Positivism is a philosophy that states that the only authentic knowledge is scientific knowledge, and that such knowledge can only come from positive affirmation of theories through strict scientific method.

 The above idea can be applied to the human psychology and social life Then, these discipline can be formulated as the

“Social Science”.

 Once the knowledge of social science is established, it is

possible to control and regulate the individual behavior and social collective movement Like the natural scientists who contribute to solving the practical engineering problems, the social scientists can also recognize the social problems and

Process of travel demand forecast

Process of travel demand

forecast

1. Identification of the travel demand

which will be forecasted.

2. Setting the target area and year.

3. Definition of the network and zoning

system.

4. Survey: data collection

5. Travel demand modeling

Travel demand forecast

1 Identification of travel

demand

 First, we should clarify the purpose/context of the

travel demand forecast

 Which transportation policy/project should be evaluated?

 How do we evaluate the policy/project with the forecasted demand?

 Then, the travel demand that will be forecasted can

be identified based on the policy purpose/context

 An analytical approach can be fixed after the travel demand which should be forecasted is specified

 Example The goal is to evaluate the introduction of new public transit service Then, the transit demand should be forecasted To do so, we need the modeling of the modal choice.

2 Setting the target area

 Geographical range which should be covered by the analysis is dependent on the goal/context of analysis

Case 1: Demand analysis of the

visitors to the Tokyo Disneyland Case 2: Demand analysis of visitors to a large-scale shopping mall in Tokyo

2 Setting the target year

 The target year of the travel demand forecast depends on the context of the transportation project/policy

土浦市

成田市 柏市

川崎市

富津市

大宮市 青梅市

八王子市

厚木市

浦和市

横浜市 木更津市

核都市広域幹線道路 東京外かく環状道路 首都圏中央連絡自動車道

北千葉道路 京葉道路

東関東自動車道水戸線

東 貫 車

常 動

東 道

新 上 路

関 動

10号線

中

線

第二 湾岸

東

クアラ

第

東京湾岸道路

千 状

横 須 路

東 速

東 動 館

東京 道路

川 道

第二東名高速道路 中央自動車道路

横 道

（平成９年度末現在）

高速自動車国道の事業中区間には 整備計画区間を含む

凡 　　 　　　　例

計　　画　　中

供　　用　　中 新五計内供用予定

Case 2: Urban master transport investment plan: 20-40 yearsCase 1: A single urban

transit project : 10-15 years

3 Definition of zoning system

 Zoning system is used to

aggregate the individual

households and premises

into manageable chunks for

modeling purposes.

 The main two dimensions of

a zoning system are the

number of zones and their

size.

 It has been common practice

in the past to develop a

zoning system specifically for

each study and

decision- Zones are represented in the travel demand models as if all their attributes and

properties were concentrated in a single point called the zone centroid.

zone centroid

A list of zoning criteria

1. Zoning size must be such that the aggregation

error caused by the assumption that all activities are concentrated at the centroid is not too large

2. The zoning system must be compatible with other

administrative divisions, particularly with census zones

3. Zones should be as homogeneous as possible in

their land use and/or population composition

4. Zone boundaries must be compatible with cordons

and screen lines and with those of previous zoning systems

5. The shape of the zones should allow an easy

determination of their centroid connectors

6. Zones do not have to be of equal size

Example of zoning system

 208 zones in Japan This is often used

in the national travel demand forecast.

model the network as a

directed graph i.e a

system of nodes and

links joining them

 Most nodes are taken to

represent junctions and

the links are

 However, there is a problem of economy versus realism which forces the modeler to select some links for exclusion

Example of transportation

network

 Road network used in the travel demand forecast in the context

of urban railway master plan (Ministry of Transport, 2000)

4 Survey: data collection

 In many urban areas,

travel survey data

plays an important

role to portray a rich

picture of the existing

methodology are of-art design, sampling and analysis, cost-

state-effectiveness and reliability for prediction over the medium to

long term

Ideal data collection method

 Collection of stage-based trip data, ensuring that

analyses can relate specific modes to specific

locations/times of day/trip lengths, etc

 Inclusion of all modes of travel

 Measurements of highly disaggregated levels of trip purposes

 Coverage of the broadest possible time period

 Collection of data from all members of the household

 High-quality data robust enough to be used even at a disaggregate level

 Integrated data collection systems incorporating

Types of travel survey

 Household survey: trips made by all household

Questionnaire survey

possible response rate and to minimize response bias.

to collect the data.

more appropriate in districts where people

are used to “filling in” forms or where

Design of questionnaire survey

 The order of the

questions normally

seeks to minimize the

respondent’s resistance

to answering them

 The difficult questions

are usually put at the

end.

 The survey is divided

into two parts: (1)

personal and household

 Sample size can be determined based on the sampling theory

5 Travel demand Modeling

Model specification

system to be modeled with a simple

structure?

forms or does the problem require more

complex non-linear functions?

and how should they enter a given model?

Model calibration/estimation

Model calibration and model estimation have a different meaning in the transportation field

 Calibrating a model requires choosing its parameters

in order to optimize one or more goodness-of-fit

measures which are a function of the observed data

 Model estimation involves the values of the

parameters which make the observed data more

likely under the model specification

Model validation

about the goodness-of-fit achieved between observed data and base-year prediction.

condition for a model validation Validation

requires comparing the model predictions

with information no used during the process

of model estimation, for example,

before-and-after studies.

6 Travel demand forecast

the planning or policy variables which are

used as inputs to the model.

quantified descriptions or scenarios about the future of the area of interest, usually using

forecasts from other sectors, models and

planning units.

Travel demand models

Travel demand model

Travel Demand

Analysis/Forecast Model

Models: Simplified representations of reality

Why are models used in transport planning?

(1) To predict future conditions in the absence of policy

intervention

Ex.) We can assess how much conditions will deteriorate.(2) To predict future conditions on the assumption that the specified policies are implemented

Ex.) We can assess the benefit

(3) To test the performance of a given policy

Ex.) We can check its “robustness”

FUNCTIONS

What is a model?

economic variables

Socio- Service (LOS) variables

Level-of-Travel demand models

Travel demand

Structure of travel demand

model

Other variables

GDP

GRDP

dan so

rong, congestion,

Desirable features of a model

(1) Accuracy and precision

 What level of accuracy is really required is determined by the context in which the model is used.

(2) Economy in data and computing resources

 What data already exist and at what cost additional data could be

provided?

(3) Ability to produce relevant indicators at appropriate level of

dis-aggregation

(4) Ability to represent relevant processes and interactions

 The model should include a representation of any processes which may

be influenced by the policy measures being tested.

(5) Appropriate geographical spread

 The model should include the whole area in which the effect of a policy might be felt.

 Ex.) Utility maximization: classical economic theory

Behavior is a result of individual’s attempting to

maximize their net gain, or minimize their net loss

 In the transport planning theory, the concept of

“minimization of the generalized cost” is sometimes

used

Fundamental concept (2)

Aggregation of individual behavior

many individuals.

abstract totals: total flow, average speed, etc

of people, their behavior can probably best

be understood by considering the behavior of

 Psychological theory

 Attitude

 Cultural, religious custom

 Economic theory

 Utility maximization under the constraints

Types of travel demand

models (2)

 Passenger

transportation

 An individual judges

where and when to go,

which mode to use, and

how frequently to travel.

 Many decision-makers are involved in a single good transportation.

 Goods themselves do not make any decision.

 Shape, unit, and characteristics may be changed through the transportation chain.

Types of travel demand

models (3)

Aggregate model vs dis-aggregate model

 Aggregate model (ex modal share)

 The model analyzes the percentage of total population

choosing a specific transportation mode.

 Dis-aggregate model (ex Individual choice of travel mode)

 An individual choose a specific travel model with some

percentage of probability.

 If we assume a representative individual in a given population, we may derive the specific alternative’s share from the representative-individual’s probability

of choosing that alternative

 We sometimes assume a representative individual by zone or by socio-demographic

Types of travel demand

models (4)

 Static model

 The model assumes that

the travel behavior is

stable during a given

period.

 It does not consider

explicitly that the one

travel behavior at a time

influences the other

travel behavior at a later

time.

 Dynamic model

 The model describes the time-series variation of travel behavior.

 The differential equation approach is applicable to the dynamic model.

 However, as the dynamic change of travel behavior

is too complex, the microsimulation approach

AVENUE (developed by Professor Masao Kuwahara,

University of Tokyo): Traffic flow and parking simulation around a shopping mall

Traffic microsimulation models

increasingly popular as a traffic analysis tool used

in transportation analyses.

 One reason for this increase is the need to model and analyze the operation of complex transportation

systems under congested conditions

 Examples: NETSIM (FHWA, US), PARAMICS (DoT, UK), AIMSUM (Polytechnic University of Catalunya, Spain)

 The analyst is assumed

to be able to observe the

data without error and to

collect the perfect data

covering all factors which

affect the traveler’s

Types of travel demand

models (6)

 Single market model

 The traditional travel

demand model covers

only transportation

market.

 The factor relating to the

other market than

transportation market is

regarded as the

exogenous variables.

 The single-market model

assumes the partial

equilibrium.

 Multi-market model

 The model takes the interaction among markets into

consideration.

 Transportation market is closely related with the other market including land market, housing market, labor market.

 The multi-market model assumes the general equilibrium.

Types of travel demand

models (7)

 Simultaneous model

 The model deals with the

individual travel behavior

as the simultaneous

equation models.

 An individual may decide

the series of travel

behaviors at the same

time, including whether

to travel, where and

when to go, which mode

to use, how often to

travel, and with whom to

travel.

 Sequential model

 The model divides the travel behaviors into several steps and computes the demand step by step.

 The simultaneous model requires the huge

computation time.

 Furthermore, actually, the individual often makes his/her decisions sequentially rather than simultaneously.