Non-Interactive Feasible Region Contraction Interior-Point Algorithm For Solution of Multi-Objective Optimization Problems

Abstract

This paper proposes an interior-point algorithm for solving multi-objective linear programming problems. The algorithm is non-interactive since it does not require an interaction with the decision maker during the process of solving multiobjective linear programming problems. It centered on locating a feasible interior-point each time the feasible region is contracted, thus generating a sequence of feasible interior-points that converge at the optimal compromise solution of a multi-objective linear programming problem. The convergence of the algorithm is proved to be on the boundary but not necessary at the extreme point of the constraint polytope. The algorithm was shown to perform with less number of iteration to reach the optimal compromise solution.