A Separation Method for Solving Multiobjective Set Covering Problem
Multiobjective optimization problems arise in many applications; hence, solving them efficiently is important for decision makers. A common procedure to solve such problems is to generate the exact set of non-dominated solutions, i.e., the Pareto front. However, if the problem is combinatorial, generating the exact Pareto fronts can be challenging. In this study, we focus on a multiobjective set covering problem and propose a separation method for generating its exact Pareto front. Particularly, the separation method first divides the problem into a set of sub-problems; then, generates the exact Pareto fronts of these sub-problems; and finally uses the sub-problem Pareto fronts to acquire the frontier of the original problem. We used the well-known Sequential Generation method to generate the exact Pareto fronts of the sub-problems. A numerical study demonstrates that separation method reduces the computational time required for generating the exact Pareto front compared to the direct application of the Sequential Generation method. This suggests that separation method is a promising approach for improving the computational time of other exact methods.
H. Farhangi et al., "A Separation Method for Solving Multiobjective Set Covering Problem," Proceedings of the 2016 Industrial and Systems Engineering Research Conference (2016, Anaheim, CA), pp. 2320-2325, Institute of Industrial Engineers (IIE), May 2020.
2016 Industrial and Systems Engineering Research Conference, ISERC 2016 (2016: May 21-24, Anaheim, CA)
Engineering Management and Systems Engineering
Keywords and Phrases
Decision Making; Multi-objective Optimization; Sequential Generation; Set Covering Problem
Article - Conference proceedings
© 2020 Institute of Industrial Engineers (IIE), All rights reserved.
01 May 2020