This thesis investigates the application of advanced mathematical programming techniques to complex manufacturing and distribution scheduling problems, with a particular focus on real-world industrial contexts. Modern production and logistics systems are characterized by high complexity, multiple interacting components, and the need to balance conflicting objectives such as cost efficiency, quality, and environmental sustainability. To address these challenges, this work develops integrated optimization models that combine exact methods with heuristic and matheuristic approaches. On the methodological side, the thesis adopts exact optimization techniques, including Branch-and-Cut, mainly to derive optimal solutions for small-size instances and to establish benchmark bounds for heuristic methods. For large-scale and computationally challenging problems, the focus shifts toward decomposition-oriented modeling and the design of tailored heuristics and matheuristics, including custom constructive heuristics, Adaptive Large Neighborhood Search, and Two-Phase Kernel Search, aimed at efficiently navigating complex solution spaces and producing high-quality solutions within practical computational limits. From an application perspective, the first part of the thesis focuses on manufacturing systems, with a specific emphasis on the modeling and optimization of decision-making problems in the steelmaking production process. Specifically, the contributions cover scrap blending under multiple objectives (cost, energy consumption, and quality), multi-period production planning with energy considerations aimed at cost reduction across casting sequences, and hybrid flow shop scheduling tailored to the steelmaking process. Novel mathematical formulations and solution algorithms are proposed, demonstrating significant improvements in computational performance and solution quality. The second part extends these methodologies to distribution systems, specifically last-mile delivery. A multi-objective optimization model is developed to support the integration of autonomous delivery robots with traditional vehicles, explicitly capturing trade-offs between operational efficiency and environmental impact. The proposed approaches provide actionable managerial insights and highlight the potential of advanced optimization for supporting sustainable logistics. Overall, the thesis contributes to the field of operations research by proposing innovative models and solution techniques that bridge theory and practice. It demonstrates how mathematical programming, enhanced by decomposition and heuristic strategies, can effectively address complex, multi-domain decision-making problems in modern industrial systems.
Questa tesi studia l’applicazione di tecniche avanzate di programmazione matematica a complessi problemi di schedulazione nei sistemi di produzione e distribuzione, con particolare riferimento a contesti industriali reali. I moderni sistemi produttivi e logistici sono caratterizzati da elevata complessità, molteplici componenti interagenti e dalla necessità di bilanciare obiettivi tra loro contrastanti, quali efficienza economica, qualità e sostenibilità ambientale. Per affrontare tali sfide, il lavoro sviluppa modelli di ottimizzazione integrati che combinano metodi esatti con approcci euristici e matheuristici. Dal punto di vista metodologico, la tesi adotta tecniche esatte di ottimizzazione, tra cui il Branch-and-Cut, principalmente per ottenere soluzioni ottime su istanze di piccole dimensioni e per fornire valori di riferimento utili alla valutazione delle euristiche. Per problemi di grandi dimensioni e computazionalmente complessi, l’attenzione si sposta verso strategie di modellazione ispirate alla decomposizione e alla progettazione di euristiche e matheuristiche sviluppate ad hoc, tra cui euristiche costruttive, Adaptive Large Neighborhood Search e Two-Phase Kernel Search, finalizzate a esplorare efficacemente spazi delle soluzioni complessi e a produrre soluzioni di elevata qualità in tempi computazionali contenuti. Dal punto di vista applicativo, la prima parte della tesi si concentra sui sistemi manifatturieri, con particolare attenzione alla modellazione e ottimizzazione di problemi decisionali nel processo produttivo siderurgico. In particolare, i contributi riguardano: il problema della carica del rottame in ottica multi-obiettivo (costo, consumo energetico e qualità), la pianificazione multi-periodo della produzione con considerazioni energetiche finalizzate alla riduzione dei costi relativi alle sequenze delle colate, e modelli di schedulazione di tipo hybrid flow shop adattati al processo siderurgico. Vengono proposte nuove formulazioni matematiche e algoritmi risolutivi, che dimostrano significativi miglioramenti in termini di prestazioni computazionali e qualità delle soluzioni. La seconda parte estende tali metodologie ai sistemi di distribuzione, con particolare riferimento al problema dell’ultimo miglio. Viene sviluppato un modello di ottimizzazione multi-obiettivo per supportare l’integrazione di robot autonomi per le consegne con veicoli tradizionali, catturando esplicitamente il compromesso tra efficienza operativa e impatto ambientale. Gli approcci proposti forniscono indicazioni utili a supporto delle decisioni manageriali e mettono in evidenza il potenziale delle tecniche di ottimizzazione avanzata nel promuovere una logistica sostenibile. Nel complesso, la tesi contribuisce al campo della ricerca operativa proponendo modelli e tecniche risolutive innovativi che colmano il divario tra teoria e applicazioni reali. Essa dimostra come la programmazione matematica, integrata con strategie di decomposizione ed euristiche, possa affrontare efficacemente problemi decisionali complessi e multi-dominio nei moderni sistemi industriali.
Mathematical programming methods for manufacturing and distribution scheduling problems / Tomasetti, L.. - (2026 Sep 29).
Mathematical programming methods for manufacturing and distribution scheduling problems
TOMASETTI, LORENZO
2026-09-29
Abstract
This thesis investigates the application of advanced mathematical programming techniques to complex manufacturing and distribution scheduling problems, with a particular focus on real-world industrial contexts. Modern production and logistics systems are characterized by high complexity, multiple interacting components, and the need to balance conflicting objectives such as cost efficiency, quality, and environmental sustainability. To address these challenges, this work develops integrated optimization models that combine exact methods with heuristic and matheuristic approaches. On the methodological side, the thesis adopts exact optimization techniques, including Branch-and-Cut, mainly to derive optimal solutions for small-size instances and to establish benchmark bounds for heuristic methods. For large-scale and computationally challenging problems, the focus shifts toward decomposition-oriented modeling and the design of tailored heuristics and matheuristics, including custom constructive heuristics, Adaptive Large Neighborhood Search, and Two-Phase Kernel Search, aimed at efficiently navigating complex solution spaces and producing high-quality solutions within practical computational limits. From an application perspective, the first part of the thesis focuses on manufacturing systems, with a specific emphasis on the modeling and optimization of decision-making problems in the steelmaking production process. Specifically, the contributions cover scrap blending under multiple objectives (cost, energy consumption, and quality), multi-period production planning with energy considerations aimed at cost reduction across casting sequences, and hybrid flow shop scheduling tailored to the steelmaking process. Novel mathematical formulations and solution algorithms are proposed, demonstrating significant improvements in computational performance and solution quality. The second part extends these methodologies to distribution systems, specifically last-mile delivery. A multi-objective optimization model is developed to support the integration of autonomous delivery robots with traditional vehicles, explicitly capturing trade-offs between operational efficiency and environmental impact. The proposed approaches provide actionable managerial insights and highlight the potential of advanced optimization for supporting sustainable logistics. Overall, the thesis contributes to the field of operations research by proposing innovative models and solution techniques that bridge theory and practice. It demonstrates how mathematical programming, enhanced by decomposition and heuristic strategies, can effectively address complex, multi-domain decision-making problems in modern industrial systems.| File | Dimensione | Formato | |
|---|---|---|---|
|
PHD_thesis_tomasetti_lorenzo.pdf
accesso aperto
Descrizione: Mathematical programming methods for manufacturing and distribution scheduling problems
Tipologia:
Tesi di dottorato
Dimensione
29.47 MB
Formato
Adobe PDF
|
29.47 MB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.


