^{}Department of Mathematics, Karaj Branch, Islamic Azad University, Karaj, Iran.

Abstract

In this paper, we propose the least-squares method for computing the positive
solution of a m n fully fuzzy linear system (FFLS) of equations, where m > n, based on
Kaman's arithmetic operations on fuzzy numbers that introduced in [18]. First, we consider
all elements of coecient matrix are non-negative or non-positive. Also, we obtain 1-cut of the
fuzzy number vector solution of the non-square FFLS of equations by using pseudoinverse.
If 1-cuts vector is non-negative, we solve constrained least squares problem for computing
left and right spreads. Then, in the special case, we consider 0 is belong to the support of
some elements of coecient matrix and solve three overdetermined linear systems and if the
solutions of these systems held in non-negative fuzzy solutions then we compute the solution
of the non-square FFLS of equations. Else, we solve constrained least squares problem for
obtaining an approximated non-negative fuzzy solution. Finally, we illustrate the eciency
of the proposed method by solving some numerical examples.