Trip generation

The first phase of transportation planning process deals with surveys, data collection and
inventory. The next phase is the analysis of the data so collected and building models to
describe the mathematical relationship that can be discerned in the trip making behaviour.
The analysis and model building phase starts with the step commonly known as trip
generation
Trip generation is a general term used in the transportation planning process to cover the field
of calculating the number f trip ends in a given area. The objective of the trip generation stage
is to understand the reasons behind the trip making behaviour and to produce mathematical
relationships to synthesise the trip-making pattern on the basis of the observed trips, land-use
data and household characteristics.
Trip
Trip is a one-way person movement by a mechanised mode of transport, having two trip ends,
an origin and destination. The first activity in travel-demand forecasting is to identify the
various trip types important to a particular transport-planning study. The trip types studied in
a particular area depend on the types of transport-planning issues to be resolved. The first
level of trip classification used normally is a broad grouping into home-based and non-homebased trips.
Home Based and Non-Home Based trips
Home-based trips are those trips that have one trip end at a household. Typical home-based
trips are the journey to work, shopping, and school. Examples of non-home-based trips are
trips between work and shop and business trips between two places of employment. The
amount of home-based trips varies from 80 to 90 percent of total travel.
Non- Home-based trips are those trips that do not have either the origin or destination end at
a household Examples of non-home-based trips are trips between work and shop and business
trips between two places of employment.
The above definitions are further classified by the following examples. Consider a trip from
home to work and the return trip from work to home. Both these trips are home based trips,
because one end of the trip is at home. Both these trips are considered to have been generated
at the home zone and attracted to the work zone. We thus have two work-purpose trip end
generations in the home zone and two work- purpose attractions in the work zone. Consider
another example of the trips from the place of work to shopping and return to the same place
of work, as it is usual during the lunch recess.

Both these trips are non-home based, because neither end of the trip is the home of the person
making trip. Both these trips are considered to have been generated at the work zone and
attracted to the shop zone. We thus have two shopping- purpose trip end generated in the
work zone and two shopping-purpose attractions in the shopping zone.
By the above definitions, total number of trip generations in any are should be equal to the
total number of attractions.
Trip purpose
Trips are made from different purposes and a classification of trips by purpose is necessary.
The following are some of the important classes of trip purpose
1. Work
2. School
3. Business
4. Social or recreational, sports
5. Others
Factors governing trip generation and attraction rates
A number of factors govern the trip generation rates. These are discussed below.
1. Income- obviously, family income which represents its ability to pay for a journey
affects the number of trips generated by a household. A general trend is that the higher
the income, higher is the trip generation rate
2. Car ownership- A car represents easy mobility, and hence a car owning household will
generate more trips that a non-car-owning household. By the same reasoning, the
more cars there are in a household, the more the number of trips generated. Of course,
number of cars owned is itself related to the income of the family, which has been
listed earlier as a factor.
3. Family size and composition. The bigger the family, the more trips there are likely to
be generated. Apart from the size, the composition of the family itself is important.
For instance, if both the husband and wife are employed. If there are many school
going children, the number of school-purpose trips will be large. If some of the
children are grown up and are employed, the number of work- purpose trips will
increase. The age structure of the family also governs the trip rates. Old persons are
not expected to generate as many trips as younger ones. The occupation of the family
is also known to influence the travel pattern.
4. Land use characteristics. Different land uses produce different trip rates. For example,
a residential area with a higher density of dwellings can produce more trips than one

with a low density of dwellings of the affluent society, which may produce a large
number of private car trips. The rateable value of the dwelling and the type of the
dwelling units affect the trip generation rates. The most important assumption made in
transportation planning is that the amount of travel is dependent on land-use.
5. Distance of the zone from the town centre- the distance of the zone from the town
centre is an important determinant of the amount of travel that people might like to
make to the town centre. The farther the town centre, the less the number if trips are
likely to be.
6. Accessibility of public transport system and its efficiency- the accessibility to a public
transport system and its efficiency determine to some extent the desire of person to
make trips. An easily accessible and efficient public transport system generates more
trips.
7. Employment opportunities, floor space in the industrial and shopping units and
offices, sales figures in shops etc.- the employment potentiality of an industrial or
shopping unit or an office establishment directly governs the trip attraction rate.
Similarly another factor to which the trip attraction rate can be related is the floor
space in the premises of industries, shops and offices.
Multiple liner regression analysis
Multiple linear regression analysis is a well- known statistical technique for fitting
mathematical relationships between dependent and independent variables. This technique has
been exploited fruitfully in a number of transportation planning studies carried out so far and
has become a very powerful tool in the hands of a transport planner. In the case of trip
generation equations the dependent variable is the number of trips and the independent
variables are the various measurable factors that influence trip generation. These independent
variables are the land-use and socio economic characteristics dicussed earlier. The general
form of the equation obtained is
Yp=a1X1+a1X2+a3X3+.anXn+U
Yp= number of trips for specified purpose p
X1, X2, X3,., Xn=independent variables relating to for example, land-use, socio economic
factors etc.
a1, a2, a3,, an= Coefficients of the respective independent variables X 1, X2, X3,., Xn,
obtained by linear regression analysis
U= Distribution term, which is a constant and representing that portion of the value of Y p not
explained by the independent variables.
The equation of the above form is developed from the present day data pertaining to
independent variables and dependent variables and the dependent variables, using statistical
techniques of least squares fitting. The equation thus developed is used for determining the
future values of trips, knowing the estimated future values of the independent variables.
The statistical theory of Multi linear regression analysis is based on the following important
assumptions
1. All the variables are independent of each other
2. All the variables are normally distributed
3. All the variables are continuous
4. A liner relationship exists between the dependent variable and the independent
variable:

5. Influence of independent variable is additive that is the inclusion of each variable in

the equation contributes a distinct portion of trip numbers.
It is difficult to ensure that the above basic assumptions are satisfied in most of the trip
generation studies. Firstly, the so called independent variables in the regression equations are
not truly independent of each other, and some sort of correlation normally exists among them.
The variables such as car-ownership, family income, residential density etc. are all interrelated to certain extent. Secondly, many of these variables are, strictly speaking, not
normally distributed. Finally, some of them are not continuous variables, an example being
the car ownership. The number of cars owned by a family can only be a discrete variable.
Aggregated and disaggregated analysis
Multiple linear regression analysis of two types:
1. Aggregated, or zonal least-square regression, where each traffic zone is treated as one
observation.
2. Disaggregated, or Household least-square regression, where each household is treated
as an observation.
The aggregated analysis which is most widely used, is based on the assumption that
contiguous households exhibit a certain amount of similarity in travel characteristics. This
assumption allows the data in a zone to be grouped and the mean value of the independent
variable used in further calculations.
Some of the disadvantages associated with the aggregated models are.
1. The analysis masks the variations in the data
2. The data is inefficiently utilised
3. The zonal sample mean is not necessarily a reliable estimate of the population mean
4. The model is dependent on the type of zoning system adopted.
5. The method is based on an important assumption that zones are to a large extent
homogenous with respect to travel and socio-economic characteristics. While care is
taken normally to select the zone boundaries to fulfil the above assumption, the
variations within the zone can sometimes be large.
6. In the process of making zones as small as possible to make them truly homogeneous,
the planner increases the complexity of analysis, notably trip distribution and
assignment.
Disaggregated analysis, though not so widely used, treats each household as an observation.
In this process, all the normous amount of data is used more efficiently resulting in a more
meaningful description of the characteristics. As compared to aggregated analysis,
disaggregated analysis produces better results and is considered more likely to be stable over
time and to provide more reliable future estimates
Criteria for Evaluation of regression Equations
The following criteria should generally be applied in evaluating and selecting a regression
equation
1. The multiple correlation coefficient should have a value at least 0.75 or even higher. A
value close to 1.0 shows a very good correlation
2. The standard error of the estimate of the dependent variable should be sufficiently
small.

3. The F test should be carried out to examine evidence of the degree of certainty that a
meaning full relationship exists between the dependent and independent variables.
4. The equation should have accuracy, validity, simplicity and sharpness and constancy.
Disadvantages of multiple-linear regression analysis technique
Some of the disadvantages associated with the assumptions made in the linear regression
technique have already been discussed earlier. In addition, the following points deserve
mention
1. The equation derived is purely empirical in nature and fails to establish a meaningful
relationship between the dependent and independent variables.
2. The technique is based on the premise that the regression coefficients initially
established will still remain unchanged in the future and can be used in the regression
equation for predicting future travel. How far the prediction is valid in future is a
main question
3. Difficulties arise in evaluation the effect of statistical problems relating to nonlinearity of the response surface and high correlation amongst the explanatory
variables.
Category analysis
Category analysis or cross classification technique is a method developed by wotton and pick
and has been used in some transportation studies in U.K. it is based on determining the
average response or average value of the dependent variable for a certain defined categories,
each dimension in the matrix reoresenting one independent variable. The independednt
variables themselves are classified into a definite number of discrete class intervals.
Assumptions
The technique is based on the following assumptions
1. The household is a fundamental unit in the trip generation process, and most journeys
begin or end in response to the requirement of the family
2. The trips generated by the household depends upon the characteristics of that
household and its location relative to its required facilities such as shops, school and
work place.
3. D with one set of household with one set of characteristics generates different trips
from the household with other set of characteristics.
4. Only three factors are of prime importance in affecting the amount of travel a
household produces: car-ownership, income and household structure.
5. Within each of the above three factors, a limited number of ranges can be established
so as to describe the trip-generating capacity of the household by a limited number of
categories.
6. Trip generation rates are stable over a time so long as factors external to the
household are the same as when the trips were first measured.
Categorization of households
As stated above, households are classified on the basis of three factors, viz., car ownership,
income and household structure. These are then classified into different ranges as indicated
below.
1. Car ownership- 3 levels, (0 car, 1 car and more than 1 car)
2. Disposable income- 6 classes:
(1) < 500 p a
(2) 500-1000 p a

(3) 1000-1500 pa
(4) 1500-2000 p a
(5) 2000-2500 pa
(6) >2500 p a
3. Household structure
i) No employed residents and one non employed adult
ii) No employed residents and 2 or more non employed adult
iii) one employed residents and one or less non employed adult
iv) one employed residents and 2 or more non employed adult
v) two employed residents and one or less non employed adult
vi) two employed residents and two or more non employed adult
Critical appraisal of the category analysis technique
The advantages that have been claimed for the technique are:
1) The whole concept of household trip making is simplified in this technique. The
technique categorises the household according to certain socio economic characteristics
and this appears rational
2) Unlike regression analysis technique, no mathematical relationship is derived between
trip making and household characteristics. This takes away many of the statistical
drawbacks of the regression analysis
3) since data from the census can be used directly, it saves considerable effort, time and
money spent on home-interview survey
4) the computations are relatively simpler
5) since disaggregate data are used, the technique simulates human behaviour more
realistically than the zonal aggregation process normally employed in regression analysis
The following are some of the disadvantages of the technique
1) It is difficult to test the significance of the various explanatory variables.
2) The technique normally makes use of the studies carried out in the past elsewhere, with
broad corrections
3) In the analysis it is assumed that income and car ownership increases in the future. The
categories of higher incomes and higher car ownership are, however, the ones which are
the least represented in the base year. Moreover, they are the ones most likely to be used
for the future estimates of trip generation
4) New variables cannot be introduced at the future date
5) Large samples are needed to assign trip rates to any one category.