A multi-objective sustainable closed-loop supply chain network problem solved by an exact method and heuristic algorithms
Date Issued
2026-06
Author(s)
Rodríguez Escoto, Joel Novi
Advisor(s)
Olivares Benítez, Elías
Nucamendi Guillén, Samuel Moisés
Type
text::thesis::doctoral thesis
Abstract
In this thesis, we present a multi-objective model for a sustainable closed-loop
supply chain network featuring hybrid facilities. The model incorporates circularity principles, sustainability factors, vehicle emissions, and the obnoxiousness of
hybrid facilities across five forward echelons, two reverse echelons, and three lo-
cation decisions. This structure leads to exponential growth in complexity across
decision levels.
The literature review reveals a key gap: classical exact and matheuristic al-
gorithms remain underexplored for multi-objective sustainable closed-loop supply
chain problems. In this context, we pioneer the application of the well-known
methods to this problem, an improved augmented ε-constraint method, a ker-
nel search matheuristic, and a hybrid non-dominated sorting genetic algorithm
II. Using a synthetically generated dataset based on supply chain parameters, we
evaluate these methods with standard Pareto front quality metrics.
Sensitivity analysis reveals the model’s behavior under varying conditions. A
comparison of the solution methods demonstrates their performance. Crucially,
the kernel search matheuristic outperforms both the classical exact method (72%
faster) and the metaheuristic (35% faster) in CPU time, delivering high-quality
solutions in the shortest time.
supply chain network featuring hybrid facilities. The model incorporates circularity principles, sustainability factors, vehicle emissions, and the obnoxiousness of
hybrid facilities across five forward echelons, two reverse echelons, and three lo-
cation decisions. This structure leads to exponential growth in complexity across
decision levels.
The literature review reveals a key gap: classical exact and matheuristic al-
gorithms remain underexplored for multi-objective sustainable closed-loop supply
chain problems. In this context, we pioneer the application of the well-known
methods to this problem, an improved augmented ε-constraint method, a ker-
nel search matheuristic, and a hybrid non-dominated sorting genetic algorithm
II. Using a synthetically generated dataset based on supply chain parameters, we
evaluate these methods with standard Pareto front quality metrics.
Sensitivity analysis reveals the model’s behavior under varying conditions. A
comparison of the solution methods demonstrates their performance. Crucially,
the kernel search matheuristic outperforms both the classical exact method (72%
faster) and the metaheuristic (35% faster) in CPU time, delivering high-quality
solutions in the shortest time.
Subjects
License
Acceso Abierto
URL License
How to cite
Rodríguez Escoto, J. N. (2026). A multi-objective sustainable closed-loop supply chain network problem solved by an exact method and heuristic algorithms. (Tesis de Doctorado). Universidad Panamericana.
Table of contents
Capítulo 1. Introduction -- Capítulo 2. Literature review -- Capítulo 3. Methodology -- Capítulo 4. Results and discussion -- Capítulo 5. Conclusions