Technical University of Crete (TUC)
Εξελικτικοί αλγόριθμοι για στοχαστικά προβλήματα διαχείρισης της εφοδιαστικής αλυσίδας
Abstract
dc:descriptionThis doctoral dissertation investigates one of the most significant problems in supply chain management, namely the Permutation Flowshop Scheduling Problem. More specifically, the problem arises in the organization of production processes, where a set of jobs must be processed on a set of machines, with the objective of determining the job sequence that optimizes a predefined performance criterion. To address this problem, a unified hybrid optimization framework based on metaheuristic algorithms was developed. In particular, nine metaheuristic algorithms were selected, each of which was combined with local improvement techniques, namely Variable Neighborhood Search and Path Relinking, as well as with a Randomized NEH algorithm for the generation of the initial solution population. The nine resulting hybrid methods were compared in terms of their performance on the classical problem, and the best-performing method was identified. The computational findings were further validated through statistical testing. The efficiency of a production schedule, however, is not determined solely by completion time, but also by additional factors such as energy consumption and operating cost. Within this context, two energy-efficient variants of the classical problem are proposed, incorporating both energy-related objectives and parameters affecting machine operation, such as processing speed. The first variant generates compromise solutions between total completion time and total energy consumption. A compromise objective function is introduced in which appropriate weighting factors assign equal relative importance to the two objectives, thereby enabling a balanced trade off between time and energy efficiency. The problem is solved using hybrid methods that were previously examined for the classical problem and proved effective in generating high-quality solutions. The selected methods are subsequently compared in order to identify the most competitive results. In the second variant, the problem is formulated as a multi-objective optimization problem involving three objectives: total completion time, total energy consumption, and the peak power of the production system. The simultaneous consideration of these objectives enables the generation of more balanced and flexible production schedules by reducing production duration, energy consumption, and peak power demand. A further and particularly important contribution of this dissertation is the incorporation of uncertainty into the two energy-efficient variants. In this way, a more comprehensive and realistic modeling framework is established for the generation of production schedules that simultaneously consider temporal, energy-related, and economic aspects. In the first stochastic variant, jobs are classified into three categories according to their level of difficulty, while processing times are modified through stochastic multipliers associated with the class to which each job belongs. The problem is solved using the proposed compromise function, and three stochastic metaheuristic methods are developed for this purpose. The resulting schedules therefore account for uncertainty while maintaining a balanced trade-off between total completion time and energy consumption. In the second stochastic, multi-objective variant, uncertainty is introduced through electricity prices. Three optimization objectives are considered: total completion time, total energy consumption, and expected electricity cost. Electricity prices are derived from historical data provided by the official Day-Ahead Market for Greece, while five different market fluctuation scenarios are additionally generated. Three multi-objective stochastic metaheuristic algorithms are proposed for solving the problem. These methods generate a set of non-dominated solutions that are evaluated under fifty different electricity price scenarios. The final outcome is a set of nondominated solutions representing different trade-offs among completion time, energy consumption, and expected electricity cost.
Degree
thesis:*- Grantor dc:publisher
- Technical University of Crete (TUC)
- Year dc:date
- 2026
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Tsiftsoglou, Maria
- Τσιφτσόγλου, Μαρία
Subjects
dc:subject × 24- Χρονοπρογραμματισμός παραγωγής
- Πρόβλημα Χρονοπρογραμματισμού Γραμμής Ροής με Σταθερή Διάταξη Εργασιών
- Ενεργειακά αποδοτικός χρονοπρογραμματισμός
- Μεθευρετικοί Aλγόριθμοι
- Υβριδικοί Aλγόριθμοι
- Πολυαντικειμενική Bελτιστοποίηση
- Στοχαστικά Προβλήματα
- Εφοδιαστική αλυσίδα
- Στοχαστικοί μεθευρετικοί αλγόριθμοι
- Permutation Flowshop Scheduling Problem
- Energy-efficient scheduling
- Multi-objective optimization
- Stochastic supply chain management problems
- Metaheuristic algorithms
- Hybrid metaheuristics
- Stochastic problems
- Stochastic algorithms
- Supply chain management
- Άλλες Επιστήμες Μηχανικού και Τεχνολογίες
- Επιστήμες Μηχανικού και Τεχνολογία
- Οργάνωση παραγωγής και Μηχανική των κατεργασιών
- Other Engineering and Technologies
- Engineering and Technology
- Industrial and Manufacturing Engineering
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/62839