Geography Reference
In-Depth Information
8.2
Representation of Urban Road Network in Multiple
Granularities
To provide an explicit representation of the relationships between roads, we use the
graph model to represent the urban road network in this paper. A graph is defined as
G
D
(
N
,
E
) in graph theory, where
N
and
E
are two disjoint finite sets.
N
is the set
of all nodes in
G
,and
E
is the set of all edges connecting two elements in
N
. Based
on this conception, we use three types of graph to represent urban road networks in
different granularity.
8.2.1
Segment Based Dual Graph
In traditional GIS network analysis and navigation system, the road networks are
always organized as primal graph: the nodes represent the junctions while the road
segment between two junctions is tread as the edge (Porta et al.
2006a
). This data
structure is from the geographic view and easy to understand. For example, the way
finding algorithms in navigation systems are based on the segment-based primal
graph. So guidance to users are always like “Go 0.4 miles and turn right to **
Avenue” which is clear and easy to follow for users.
However, further relationship between road segments such as “Does certain road
segments have more neighbors than others?” or “Do neighbors of one road segments
tend to connect with each other?” are difficult to provide directly from the primal
graph. To overcome this shortage, dual graph, where road segments are nodes and
the junction between road segments map to edges between nodes is constructed to
represent the relationship between road segments (Porta et al.
2006b
). In Fig.
8.1
we present a simple example of a piece of urban road network mapped to such dual
graph.
Figure
8.1
a shows a sketch map of initial urban road network where nodes are
intersections and edges are road segments between intersections. It can provide
both location and geometric shape for road segments. In Fig.
8.1
b, road segments
are shown as nodes and the direct connections between segments are shown as
edges. It can provide more information about the connections than Fig.
8.1
aand
these information is valuable for robustness analysis. In the following part of
this article, “road-segment based road network” refers to road-segment based dual
graph.
Road segments in Fig.
8.1
show variation in the network structure. This variation
can be used in robustness analysis. For example, if segment 1 is attacked, the
rest of the network can still be functional. But if segment 5 is attacked, vehicles
from segment 1-4 to segment 7-10 can only take segment 6, which can put great
pressure on segment 6. If Segment 6 is malfunctioned too, the network is split into
two components and becomes malfunction. This suggests that attack on different
segment can cause different damage.
Search WWH ::
Custom Search