Vous êtes sur la page 1sur 3

QMM Assignment

Submitted by : Kushagra Agarwal


Registration No : 2016SMT6713

Minimum Spanning Tree Linear Programming


Formulation
A minimum spanning tree is a subset of edges of a connected graph that
connects all the vertices together without any cycles and minimum possible
edge weight.

In this problem, we will be given a graph and a function on the edges, we


want to find out a spanning tree that fulfills the objective of minimizing the
overall function value under certain constraints. We have to select a subset
of the edges that form a spanning tree.
Let

x ij

represents the edge from node i to j.


x ij

=1
=0

Edge is a part of spanning tree.


Otherwise.

Let the function on edges be Cost (

c ij

). Then our objective is to minimize

the total edge cost:


Z = c ij x ij
Following are the constraints:
1. There should be (n-1) edges in tree.
2. There should be no cycles in tree.

Let us explain this with the help of an example:

Let us develop the constraints for this problem. As before, we assign a


variable to each edge of the graph, and, if the variable xij is 1 then the edge
from node i to node j is in the tree. To find the minimum spanning tree we
minimize the sum of the variable-weight products.

First constraint: We require exactly 5 edges in the tree. Therefore,


x ab+ x ac + x ae + x bc + x bd + x cd + xce + x cf + x df + x ef =5
The above equation is not sufficient as we can have the feasible solution by
selecting smallest five edges but that would not span the graph. To achieve
this we have to ensure that atleast one edge corresponding to each node
must be there in the tree. This requirement results in following constraints:
x ab+ x ac + x ae=1

[Vertex a]

x ab+ x bc + x bd=1

[Vertex b]

x ac +x bc + x cd +x ce +x cf =1

[Vertex c]

x bd + x cd + x df =1

[Vertex d]

x ae + x ce+ x ef =1

[Vertex e]

x cf + x df + x ef =1

[Vertex f]

Second constraint: Taking all sub-graphs in which there is a possibility of


formation of cycles. For example, sub-graph containing nodes {a,b,c} might
have a cycle of 3 therefore,
x ab + x ac + x bc 2
Similarly for sub-graph {a,b,c,d}:
x ab + x ac + x bc + x bd + x cd 3
We can write similar constraints for the various cycles possible. There are 26
such constraints.
Now, we can apply Simplex method to solve the above equations and obtain
an optimal solution.

Vous aimerez peut-être aussi