A.A hosseinzadeh - Department of Mathematics, Lahijan Branch, Islamic Azad University, Lahiaj, Iran
A Ghane –Kanafi - Department of Mathematics, Lahijan Branch, Islamic Azad University, Lahiaj, Iran

In most of real world problems data suffer from being inaccurate thus considering an interval for perturbation of data seems to be an alternative. There exist many papers in literature about Interval linear programming (ILP) problems dealing with nonnegative variables. Here the general form of ILP is considered where all the parameters and variables are considered to be interval and moreover variables are unrestricted in sign. This issue due to high complexity has not been dealt with in literature but this is the case in most of the real world problems. In this paper a new method presented in order to obtain fully interval linear programming problems (FILP_P) using a nonlinear programming problem (NLP). Furthermore, in order to demonstrate how the proposed method works two numerical examples are illustrated.

Interval linear system, Interval number vector, Linear programming, Ranking interval