Opsearch, Vol. 34, No. 1, 1997
0030-3887/95 $ 2.00 + 0.00
© Operational Research Society of India
New Technique for Solving Primal
All-integer Linear Programming
Shambhu Sharma
Department of Mathematics, L.P. Shahi College, Patna, India
and
Dr Bishram Sharma
Department of Mathematics Science College, Patna, India
Abstract
This paper provides a new technique of selecting the pivot column and
the source row in primal all-integer linear programming. by which the
number of iterations has been reduced in going through the initial simplex
table to the optimal simplex table.
1. INTRODUCTION
All the existing methods to solve the integer linear programming problem
[2, 3, 4] are based on the concepts of selecting the pivot column and the
source row as given in the linear programming. None of these methods exploit
the simple and well known fact that the method of selecting the pivot column
and the source row, given in the linear programming does not give the better
improvement of the objective function.
We have searched the pivot column by minimum of the product of the
negative coefficient of the objective function and the minimum of the ratios
used in [4] for the source row. The source row automatically comes out by
this technique. However the formulation of cut and the iteration process of
our technique are the same as the other well known methods.
2. THEORETICAL FOUNDATIONS
Consider the problem
Max Xo cx
Such that Ax = b
(1 )
x 0, integers