Networks in Transportation Theory
Networks in Transportation Theory
I. Introduction
A graph >G= in the above sense consists of two things: a set V whose
elements are called >vertices= or >nodes=; and a set E of unordered
pairs of distinct vertices, called >edges=. It is commonly denoted as
G(V, E). In the example of a simple graph G below, the set of vertices
(points) V={a,b,c,d,e} and the set of edges(lines) E={ab,ac,bc,cd,de}.
There are some names used for special types of graphs, null graphs
with no edges, complete graphs with every vertex joined to every
other vertex, cycles which only join the outside of the vertices,
wheels which add a vertex at the centre (see below from Rosen, p.
448).
Complete
(Economides, p. 110.)
A) Routing
i) The one way street problem - One of the problems faced by cities
to control the flow of traffic is to ask whether it would be possible
to convert two-way streets into one-way streets without getting into
trouble i.e., ending up with some places that one can get into and
never leave. That is for every pair of locations (x and y), is it
possible to reach x from y and vice versa? Let us take a graph >G=
in the shape of a square with four sides, then let us give it a
orientation (i.e., direction to each edge of G or diagraphs going
clockwise) that is strongly connected [1]. In such a case, it is
always possible to get from one vertex to another.
(Roberts, p. 450)
To solve this problem one has to ensure that Robbins= theorem (which
states: A graph G has a strongly connected orientation if and only if
G is connected and has no bridges) is satisfied. But the theorem does
not tell us how to find an assignment for a connected bridgeless
graph. One way to do so is to use an algorithm known as A Depth
First Search Algorithm. This algorithm, however, does not tell us
whether it is efficient or inefficient in the sense of being the shortest
one way street assignment. Fred Roberts states that there is evidence
that no such algorithm exists.
B) Networks
The Concise Oxford dictionary defines it as an Aarrangement with
intersecting lines & interstices... In mathematical topology, it is
defined as ...a figure (in a plane or in space) consisting of a finite,
non-zero, number of arcs, no two of which intersect except possibly
at their end.[4] In this sense, it is more restrictive than some of the
graphs described earlier.
ii) The minimum cost flow problem The minimum cost flow
problem is finding the cheapest possible way of sending a certain
amount of flow (positive) from source s to sink t through the network
at a minimum cost (i.e., nonnegative capacity (cij ) on each arc and
nonnegative cost (aij )). The transportation problem can be viewed as
a minimum cost flow problem. Imagine that a particular commodity
is stored in n warehouses and is to be shipped to m markets. Let ai
and bi be the supply and demand at each warehouse and market,
respectively. Let aij be the cost of transportation and let total supply
equal to total demand (i.e., 3ai = 3bj ). The problem is to find a
shipping pattern that minimizes the total transportation cost (3aij xij
i=1...n, j=1...m) subject to certain constraints --- total amount of
commodity shipped from the ith warehouse is at most the amount
there (3 xij #ai, j=1...m) and total amount of commodity shipped to
the jth market is at least the amount demanded (3 xij #bj, i=1...n). Any
solution satisfying these two constraints will also satisfy 3 xij =ai,
j=1...m; and 3 xij =bj, i=1...n - since supply equals demand.
V. Concluding Remarks
Networks have always been important in transportation. They are
widespread in Canada providing everyday links from home to work,
life lines to diverse regions together with corridors for our export and
import trade. Our economic futures are more closely tied to the
sustainablility of our transportation networks than we might care to
admit. They define the key to success in todays dynamic
marketplace. They are a strategic tool to gain competitive advantage.
This is particularly true today in the context of globalization.
Bibliography
1. Barabasi, A., Linked: the new science of networks, Perseus Publication, 2002.
2. Barabasi, A. L. and Albert, R., Emergence of scaling in random networks,
Science, Volume 286, 1999, pp. 509-512.
3. Cervero, R. and Hansen, M., Induced travel demand and induced road
investment: a simultaneous equation analysis, Journal of Transport Economic
Policy, 2002, Volume 35, pp. 469-490.
4. Economides, N., The economics of networks, International Journal of
Industrial Organization, Volume 14, No. 6, 1996, pp. 673-699.
5. Economides, N., Competition Policy in Network Industries: An Introduction,
Texas A&M University.
6. Economides, N., Network externalities, complementarities, and invitations to
enter, European Journal of Political Economy, Vol. 12 (1996), pp. 211-233.
7. Feng Xie and David Levinson, Modeling the Growth of Transportation
networks: A Comprehensive Review, New Spatial Economics, Volume 9,
2009, pp. 291-307.
8. Feng Xie and David Levinson, Topological evolution of surface transportation
networks, Computers Environment and Urban Systems, Volume 33, 2009, pp.
211-223.
9. Fred Roberts, Applied Combinatorics, Rutgers University, 1984.
Endnotes
[1] To show that a graph is connected one usually employs a highly efficient procedure
known as the depth-first search procedure. Efficiency here refers to computational
complexity.
[2] See Fred Roberts, Applied Combinatorics, Rutgers University, 1984, pp. 460-461.
[3] See Wikipedia, www.wikipedia.org
[4] Arnold, B.H., Intuitive Concepts in Elementary Topology, 1963, p. 31.
[5] The others are listed in maximum flow problems in www.wikipedia.org
[6] Network Economics: An Introduction, Anna Nagurney, 2002.
[7] Feng Xie and David Levinson, Topological evolution of surface transportation
networks, Computers Environment and Urban Systems, Volume 33, 2009, p. 211; and