Scribd
Upload
39 views2 pages

Network TABLE

The document discusses several network models: minimum spanning tree, shortest path, maximum flow, and minimum cost flow. It provides details on the topology, parameters, variables, objectives, constraints, characteristics of the solution, algorithms, examples, and notes for each model. The models vary in their directed/undirected nature, sources/sinks, costs, supplies/demands, and how flow is handled. Minimum spanning tree finds the lowest cost connected network while shortest path finds the lowest cost path between a unique source and sink. Maximum flow determines the maximum amount of flow that can pass from a source to sink along multiple paths. Minimum cost flow aims to route flow at minimum cost while satisfying demands.

Uploaded by

djdj7752
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
39 views2 pages

Network TABLE

The document discusses several network models: minimum spanning tree, shortest path, maximum flow, and minimum cost flow. It provides details on the topology, parameters, variables, objectives, constraints, characteristics of the solution, algorithms, examples, and notes for each model. The models vary in their directed/undirected nature, sources/sinks, costs, supplies/demands, and how flow is handled. Minimum spanning tree finds the lowest cost connected network while shortest path finds the lowest cost path between a unique source and sink. Maximum flow determines the maximum amount of flow that can pass from a source to sink along multiple paths. Minimum cost flow aims to route flow at minimum cost while satisfying demands.

Uploaded by

djdj7752
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/ 2

Network Models

Minimum Shortest Path Maximum Flow Minimum


Spanning Tree Cost Flow

Topology
•Arcs Undirected Directed Directed Directed
•Source and None Unique source and Unique source and Multiple
sink unique sink unique sink sources/sinks
Parameters
•Supplies/ None 1 at source and sink, A variable, v, at Varied at
demands 0 elsewhere source and sink many nodes
•Costs cij cij none cij
•Lower/ none none (lij , uij ) on arcs (lij , uij )
upper bounds on arcs
Variables 1 if use an arc, 1 if use an arc, flow flow
0 otherwise 0 otherwise
Objective Minimize cost Minimize cost Maximize flow Minimize cost
P
Constraints arcs = n − 1, Flow balance at Flow balance, Flow balance,
Subtour elimination all nodes lij ≤ flow ≤ uij lij ≤ flow ≤ uij
Characteristics
of Solution
•Flow All nodes simply One path from s to t; Multiple paths from See
connected ≤ one of each arc s to t; paths contain maximum flow
and node nodes/arcs > once
•Fractional flow Does not make Does not make sense ok if bounds ok if bounds or
sense fractional supply/demand
fractional
Algorithm(s) Kruskal Dijkstra, Label Shortest augmenting Network
correcting path simplex
Examples Message passing Replace equipment Pipelines Warehouses
Sprinkler systems Staffing Open pit mining Race balance
Air duct pathways Property rental Baseball elimination Trains
Computer networks Knapsack problems Matching resources Airplanes
Data clustering Library stacks Table seating Manufacturing

1
Notes:

• All models, other than the minimum spanning tree, contain flow balance constraints and
nonnegativity.

• In shortest path and spanning tree models, the variables are restricted to be ≤ 1.

• It is NEVER necessary to impose integrality requirements on any network flow models as


long as the constraint data are integer.

• Trivial solutions occur in a shortest path model with one path from s to t with all zero-
length arcs (as long as all arc lengths are nonnegative).

• A maximum flow model has an obviously feasible solution if there exists a path from s to
t with 0 lower bounds on all the arcs in the path.

• An unbounded solution will occur in a shortest path model given a negative cost cycle.

• An unbounded solution will occur in a maximum flow model given at least one path from
s to t with all arcs in the path possessing no finite upper bound.

• Minimum cost flow models can be unbounded, infeasible, or have multiple optima, in
addition to having a unique solution. Minimum cost cycles yield unboundedness; Phase I
can be applied to find an initial feasible solution, as in LP simplex.

• Transportation models are special cases of minimum cost flow models in which the graph
is bipartite and one set of nodes is matched with another set of nodes; there are no arcs
between nodes in the same set. One set of nodes has supply, and the other contains
demand.

• Assignment models are special cases of transportation models in which all supplies and
demands are 1.

• Degeneracy occurs much more frequently in networks than in standard linear programs.

You might also like