APPLICABILITY OF EXISTING CONTINUOUS METHODS IN DETERMINING THE PARETO SET FOR NONLINEAR, MIXED-INTEGER MULTICRITERIA OPTIMIZATION PROBLEMS

This article investigates the applicability of existing techniques for generating the Pareto set in continuous problems to nonlinear multicriteria optimization problems where some of the optimization variables are discrete. Techniques such as the weighted sums approach, goal programming and Normal-Boundary Intersection are commonly used to generate points on the Pareto surface for continuous problems. These approaches are re-evaluated in the light of then- geometrical properies for mixed-integer nonlinear problems, which pose an additional challenge because the Pareto surface is not connected. It is demonstrated that the intuition we have gathered from continuous problems may not apply in the discrete setting.