Comparison of Four Goal Programming Algorithms

Abstract

A major limitation in the use of goal programming has been the lack of an efficient algorithm for model solution. Schniederjans and Kwak recently published a proposal for more efficient solution of goal programming models utilizing dual simplex procedures. A goal programming algorithm based upon that method has been coded, as well as a revised, simplex-based algorithm. These algorithms are compared in terms of accuracy and time requirements with algorithms previously presented by Lee and by Arthur and Ravindran. Solution times for a series of 12 goal programming models are presented. The dual simplex method appears to have superior computational times for models with a large proportion of positive deviational variables in the solution. The revised simplex algorithm appears more consistent in time and accuracy for general goal programming models.

Keywords:

• goal programming
• revised simplex
• dual simplex