Republic of the Philippines

Autonomous Region in Muslim Mindanao

ModuleNo.1:
Module No.4:Introduction
Trip Distribution
to Highway Engineering and Traffic Analysis

TIME FRAME: 1 Week Lecture

TRIP DISTRIBUTION

 The objective of this model is to determine from which zones these Trip Generations (I) are
generated and to which zones these trip attractions ( J ) are bound

 THE ORIGIN-DESTINATION TABLE(OD MATRIX)

DESTINATION ∑

1 2 … J … n-1 n

O 1 O1

R
2 O2
I
.. …
G

I i Tij Oi

N
.. …

n-1 On-
1

n On

∑ D1 D2 … Dj … Dn-1 Dn T
TRIP DISTRIBUTION INPUTS AND OUTPUTS

THE TRIP DISTRIBUTION MODEL

QIJ = PI (AJ FIJ KIJ)/(∑AJFIJKIJ)= PIpIJ

this is known as the gravity model

where:

FIJ = friction factor or friction of travel =1/WCIJ

AJ = trip attraction at zone J

KIJ = socioeconomic adjustment factors to incorporate factors not captured by the

PI = trip production at zone I

pIJ = probability that a trip generated by zone I will be attracted by zone J. When
multiplied by 100%, it is the percentage of trips going into zone J.

A*J = ∑ QxJ
where:

A*J = Total trip attraction

QxJ = interchange volume

EXAMPLE (APPLICATION OF THE GRAVITY MODEL )

The target year productions and relative attractiveness of the four-zone city have been estimated
to be as follows:

Zone Productions Attractiveness

1 1500 0

2 0 3

3 2600 2

4 0 5

The calibration of the gravity model for this city estimated c=2.0 and all socioeconomic
adjustment factors equal to unity. Apply the Gravity Model to estimate all target year interchange QIJ
and total target year Attractions of each zone given the target year interzonal impedances as follows:

I J 1 2 3 4

1 5 10 15 20

2 10 5 10 15

3 15 10 5 10

4 20 15 10 5
SOLUTION

 For I = 1, P1 = 1500

J AJ FIJ KIJ AJFIJKIJ pIJ QIJ

1 0 0.0400 1.0 0.0000 0.000 0

2 3 0.0100 1.0 0.0300 0.584 875

3 2 0.0044 1.0 0.0089 0.173 260

4 5 0.0025 1.0 0.0125 0.243 365

∑= 0.0514 ∑= 1.000 ∑= 1500 = P1

 For I = 3, P3 = 2600

J AJ FIJ KIJ AJFIJKIJ pIJ QIJ

1 0 0.0044 1.0 0.000 0.000 0

2 3 0.0100 1.0 0.030 0.188 488

3 2 0.0400 1.0 0.080 0.500 1300

4 5 0.0100 1.0 0.050 0.312 812

∑= 0.160 ∑= 1.000 ∑= 2600 = P3

A*1 = Q11 + Q21 + Q31 +Q41 = 0 + 0 + 0 + 0 = 0

A*2 = Q12 + Q22 + Q32 +Q42 = 875 + 0 + 488+ 0 = 1363

A*3 = Q13 + Q23 + Q33 +Q43 = 260 + 0 + 1300 + 0 = 1560

A*4 = Q14 + Q24 + Q34 +Q44 = 365 + 0 + 812 + 0 = 1177

 TRIP TABLE: Summary of the solution

I J 1 2 3 4 PI

1 0 875 260 365 1500

2 0 0 0 0 0

3 0 488 1300 812 2600

4 0 0 0 0 0

0 1363 1560 1177 4100

CALIBRATION OF THE GRAVITY MODEL

Illustration Problem:

Consider the 5-zone city shown below. Two of the zones are purely residential, and the
remaining three are purely nonresidential. The base-year interzonal impedances are specified in terms of
travel time in minutes and are shown in parentheses on the arc joining pairs of zones. The observed base-
year productions, attractiveness, and trip interchange volumes are inserted in the figure. Calibrate the
Gravity Model.

FIGURE: (a) 5-zone City

ZONE I PI AI

1 500 0

2 1000 0

3 0 2

4 0 3

5 0 5

(c) Base-year distribution

J 3 4 5

1 300 150 50

2 180 600 220

Trip length Frequency Distribution Computation

W ∑QIJ f = (2)/Sum

(2)

5 300 +600 = 900 0.60

10 150 + 180 = 330 0.22

15 50 + 220 = 270 0.18

Sum = 1500

 FIRST ITERATION:

The calculation procedure is followed by ASSUMING AN INITIAL ESTIMATE FOR “c” say
c =2.0 and first considering KIJ = 1, use the Gravity Model Formula to solve values of
interchange volumes QIJ and the frequency distribution as follows:

RESULTING INTERZONAL VOLUMES ESTIMATE OF THE FIRST ITERATION with c = 2.0

 Base Year Trip Distribution:

J 3 4 5

1 303 114 83

2 123 741 136

W ∑QIJ f = (2)/Sum

(2)

5 303 + 741 = 1044 0.70

10 114 + 123 = 237 0.16

15 83 + 136 = 219 0.14

Sum = 1500

DISCUSSION OF THE RESULT OF FIRST ITERATION

Application of the Gravity Model Formula using the assumed c=2.0 leads to the interzonal
estimates shown in tabulation above along with the calculated trip-length frequency distribution
superposed on the observed trip length frequency distribution to illustrate the discrepancy between the
two.

- the assumption that c = 2.0 is seen to overestimate the percentage of trips at low impedances and to
underestimate the percentage at high impedances.

ITERATION 2 WITH ADJUSTED FIJ - FACTORS , FIJ*

To rectify the situation of overestimated and underestimated values in ITERATION 1, the FIJ –
FACTORS in different impedance levels are adjusted. Commonly used formula is:

FIJ* = FIJ (Observed / Calculated)

Where: FIJ* = the adjusted friction factor at a given impedance WIJ

Observed = the corresponding base-year percentage of trips

Calculated = the current estimate of the percentage of trips at that impedance level.

For Example, at WIJ = 5, assumed FIJ = 1/52 = 0.04 then:

F*IJ5 = 0.04(0.60/0.70) = 0.034, then calculate F*IJ10 , and F*IJ15 and these factors are used in the
Gravity model to solve for trip distributions .
RESULTING INTERZONAL VOLUMES ESTIMATES OF THE SECOND ITERATION
WITH ADJUSTED FRICTION FACTORS
FIJ*.

Adjusted Trip Distribution

J 3 4 5

1 251 145 104

2 176 654 170

Trip-length Frequency Distribution

W ∑QIJ f = (2)/Sum

(2)

5 251 +654 = 905 0.60

10 145 + 176 = 321 0.21

15 104 + 170 = 274 0.19

Sum = 1500

DISCUSSION FOR ITERATION 2

- as seen in the Trip-length Frequency Distribution compared to the Observed , percentage values
are closer therefore it can be the last iteration, but if in case values are still far then continue the
process using adjusted Friction Factors, this time your FIJ is the adjusted factors in the second
iteration or in this iteration, Continue until values become refined.

- to determine final value of “c” is to perform Simple Linear Regression of the estimated values of
F from the observed to the last iterations made. Or maybe as a sort of simple estimation take the
average of the Friction Factors of the last iteration and solve for a single value of “c”.
Determination of the KIJ values:

- Even though the trip-length frequency distribution is now close to the observed distribution,
certain pronounced discrepancies remain at the interchange level. To adjust these, the calibration
procedure FINE TUNES the Model by introducing a set of zone-to-zone SOCIOECONOMIC
ADJUSTMENT FACTORS calculated as follows:

KIJ = RIJ{(1-XI)/(1-XIRIJ)}

Where: RIJ = the ratio of Observed to calculated QIJ

XI = the ratio of the base-year QIJ to PI, the total production of Zone I

For Example: for interchange 1-3:

R13 = 300/251 and X1 = 300/500 then calculate K13 and so on …..

KIJ Values for these Calibration