Scribd
Upload
13 views4 pages

Intervening Opportunities Model 071019

Uploaded by

ing.valenza
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
13 views4 pages

Intervening Opportunities Model 071019

Uploaded by

ing.valenza
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
You are on page 1/ 4

The Intervening Opportunities model

Consider a zone i. All zones j = 1, ..., n are in order of increasing distance


from i. Zones include also zone i, thus giving rise to a model providing also
intra-zonal trips. Thus zone j = 1 is zone i.
Assume the following notation.
Ti(j+1) , j = 0, ..., n − 1, number of trips between zone i and zone j + 1,
Oi , number of total trips originating from zone i,
Pi(j+1) , j = 0, ..., n − 1, probability of progressing from zone i up to zone j + 1
included or of not making a trip,
Aj+1 , j = 0, ..., n − 1, number of opportunities up to zone j + 1 included,
L, constant probability that a possible destination is accepted if it is consid-
ered.
We have evidently:

(1)
 
Ti(j+1) = Oi · Pi(j+1) − Pij , j = 0, ..., n − 1,
since Pi(j+1) − Pij is the probability of a trip terminating in zone j + 1. Zone
j = 0 is a ctitious zone. Pi0 is the probability of not making a trip.
The model assumption is that the probability Pi(j+1) − Pij of a trip termi-
nating in zone j + 1 is directly proportional to the number of opportunities
(Aj+1 − Aj ) in zone j + 1. Therefore we can write:

Pi(j+1) − Pij = L · (1 − Pij ) · (Aj+1 − Aj ) , j = 0, ..., n − 1, (2)


because, evidently, by the law of total probability:

Pi(j+1) = Pij + (1 − Pij ) · L · (Aj+1 − Aj ) , j = 0, ..., n − 1. (3)


We assume that the number of opportunities A0 in j = 0 is zero.
Transferring to continuous variations (i.e. to dierentials) we get:

dp = L · (1 − P ) · dA. (4)
If we integrate the ordinary dierential eqn (4) we obtain:
dP
= L · dA, (5)
1−P

1
ln (1 − P ) = −L · A + k, (6)
and therefore

P = 1 − k · e−L·A . (7)
In zonal notation this becomes:

Pij = 1 − ki · e−L·Aj , j = 0, ..., n. (8)


Substituting eqns (8) into eqn (1) we obtain the classical expression of the
intervening opportunities model:

(9)
 
Ti(j+1) = ki · Oi · e−L·Aj − e−L·Aj+1 , j = 0, ..., n − 1.
The meaning of the constant of integration ki can be derived from eqn (8)
written for j = 0:

Pi0 = 1 − ki , (10)
which gives:

ki = 1 − Pi0 , (11)
i.e. ki is the probability of making a trip.
Notice that the intervening opportunity model considers all events: not mak-
ing a trip, progressing up to zone 1 included, ..., progressing up to zone n
included, progressing beyond zone n. The probability of progressing beyond
zone n is given by 1 − (Pin − Pi0 ) .
In practice, the numerical value of the constant of integration ki is usually
obtained by imposing the origin constraint on the number of trips originating
from i: n
(12)
X
Tij = Oi .
j=1

Substituting eqns (9) into eqn (12):


Oi = ki · Oi · (1 − e−Li ·An ), (13)
and solving for ki we obtain:

2
1
ki = . (14)
1 − e−Li ·An
Notice that this is equivalent to imposing that the sum, extended to the
destinations j = 1, ..., n, of the model probabilities of a trip terminating in j
equals unity:
(15)
 
(Pi1 − Pi0 ) + (Pi2 − Pi1 ) + ... + Pin − Pi(n−1) = 1,
which gives:

Pin − Pi0 = 1, (16)


i.e. the probability of progressing up to n is one, and

1 − ki · e−L·Ai − 1 + ki = 1, (17)
from which the expression of ki in eqn (14) is obtained. This is, in turn,
equivalent to impose that the probability of progressing beyond n is zero,
having considered that this probability is 1 − (Pin − Pi0 ) .
The magnitude of ki depends on the unit of measurement chosen for the
opportunities Aj , j = 1, ..., n. The number of trips terminating in each zone
can be assumed as surrogate (proxy) of the number of opportunities in the
zone. However, the model is only singly constrained, not doubly constrained,
i.e. only the constraint on the trips originating from each zone is satised,
the constraint on the number of trips moving into each zone is not satised.
Li is a calibration parameter. Usually, it is determined on the basis of the
observed trip length distribution for zone i. From this it is possible to deter-
mine the observed average trip length di . Therefore, the parameter Li is the
solution of the eqn obtained by equalling the observed average trip length di
with the modelled average trip length (lengths weighted with the number of
trips):
Pn−1  
di(j+1) · e−Li ·Aj − e−Li ·Aj+1
(18)
j=0
di = .
1 − e−Li ·An

3
Used bibliographical references
Akwawua S., Pooler J.A. (2001) The development of an intervening opportu-
nities model with spatial dominance eects. Journal of Geographical Systems
3, 69-86.
Ruiter E.R. (1967) Towards a better understanding of the intervening op-
portunities model. Transportation Research 1, 47-56.
Wilson A.G. (1970) Entropy in Urban and Regional Modelling. Pion, Lon-
don.
Wilson A.G., Bennett R.J. (1985) Mathematical Methods in Human Geog-
raphy and Planning. Wiley, Chichester.

You might also like