An algorithm for mixed integer linear fractional programming problems

Abstract

A new algorithm for solving mized integer linear fractional programming problem which maximizes a linear fractional objective function under the constraint of some linear inequalities is developed in the present paper. After obtaining an optimal solution of the continuous fractional problem whatever the feasible region is, we are able to construct an improved branch and bound method based on computation of penalties for solving the mized integer linear fractional problem in a finite number of iterations. An illustrative numerical erample is included.