Athens University Economics and Business (AUEB)
Optimization methods for complex vehicle routing and scheduling problems
Abstract
dc:descriptionThis PhD dissertation is interested in the development of optimization methods for solving complex and large scale Vehicle Routing and Scheduling Problems (VRSP). The main focus is given on studying the computational efficiency and effectiveness of hybrid metaheuristic algorithms. In an attempt to fill in the gaps of the literature identified earlier, the major effort will focus on the Vehicle Routing Problem with Time Windows (VRPTW) and the proposed optimization methods will be tested on several medium and large scale benchmark data sets. Attention will be also given on other variants of the VRPTW with practical and operational side-constraints namely the Open VRPTW (OVRPTW) and the Heterogeneous Fleet VRPTW (HFVRPTW) and the VRPTW with Multiple Compartments and Commodities (MCCVRPTW). Finally, driven from an industrial application we will examine the integration of the suggested methods within a novel Decision Support System that enables schedulers to tackle VRPTW problems with real life operational constraints. A hybrid Multi Start Local Search and a novel Arc Guided Evolutionary Algorithm are proposed for solving the VRPTW. For both approaches an extensive empirical study is performed and the role of key algorithmic elements is analyzed. Computational experiments demonstrate that the proposed optimization methods obtain the best overall results compared to the current state of the art. Considering the Open VRPTW a formal description is first provided followed by a greedy look-ahead construction heuristic and a hybrid Evolutionary Algorithm. All critical aspects of the problem are captured while experimental results demonstrate the competitiveness of the methods with respect to other approaches. Subsequently, the HFVRPTW is addressed. A novel two phase solution framework is developed called Reactive Variable Neighborhood Tabu Search. Computational experiments show that high quality solutions can be reached via the proposed method illustrating its efficiency and effectiveness. Next a real-life VRPTW is presented that incorporates a number of side constraints such as multiple compartments and commodities, time windows, loading restrictions and limited vehicle fleet size and composition availability. An Adaptive Memory Programming method is proposed to address all critical aspects of the problem. Various computational experiments demonstrate the applicability of the method to real life vehicle routing and scheduling applications. Finally, an industrial Decision Support System is presented that handles the efficient and effective management of waste lube oils collection and recycling operations. The main effort is to examine the integration and cooperation of the vehicle routing and scheduling methods within an Enterprise Resource Planning system. The application of the DSS to an industrial environment improved productivity and competitiveness.
Degree
thesis:*- Grantor dc:publisher
- Athens University Economics and Business (AUEB)
- Year dc:date
- 2008
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Repoussis, Panagiotis
- Ρεπούσης, Παναγιώτης
Subjects
dc:subject × 20- Διαχείριση εφοδιαστικής αλυσίδας
- Δρομολόγηση οχημάτων
- Υπολογιστική ευφυΐα
- Ευρετικοί αλγόριθμοι
- Μεταευρετικοί αλγόριθμοι
- Εξελικτικός προγραμματισμός
- Συστήματα διανομών
- Διοίκηση υπηρεσιών και επιχειρησιακών διαδικασιών
- Supply chain management
- Logistics
- Vehicle routing and scheduling
- Metaheuristics
- Heuristics
- Evolutionary computation
- Distribution management
- Service operation management
- Κοινωνικές Επιστήμες
- Οικονομικά και Επιχειρήσεις
- Social Sciences
- Economics and Business
Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/25322