A Semi-Definite Programming Approach for the Single-Row Facility Layout Problem with Multiple Objectives

Description

In this work, we combine ideas from the fields of Semi- Definite Programming (SDP) and Multi-Objective Program- ming (MOP) to compute the Pareto frontier of combinatorial optimization (CO) problems. Our focus is on the Single-Row Facility Layout Problem (SRFLP). The SRFLP is a well-known NP-hard CO problem in the field of industrial engineering. The problem consists of finding the optimal arrangement of facilities along a single row, such that the total cost is minimized. The SRFLP is frequently extended to a multi-objective optimiza- tion problem, to simultaneously consider conflicting objectives, such as flow distance, CO2 emissions or total material handling cost. We propose a novel approach that leverages the power of SDP to solve the SRFLP with multiple objectives. Specifically, our focus is to combine objective space techniques, such as the weighted sum method and the epsilon-constraint method, with SDP. Solution algorithms based on these techniques are used in a computational study to compare our novel approach with ob- jective space methods using integer programming formulations. For this comparison, we consider an adapted set of benchmark instances from the literature.

Period	25 Jul 2024
Event title	25th International Symposium on Mathematical Programming
Event type	Conference
Location	CanadaShow on map