Athens University Economics and Business (AUEB)
Models and solution algorithms for inventory routing problems
Abstract
dc:descriptionThe main objective of this thesis is to propose a hybrid evolutionary optimization algorithm for solving the Inventory Routing Problem (IRP). The IRP arises from the application of the Vendor Managed Inventory (VMI) concept, where the supplier (vendor) has to make inventory and routing decisions simultaneously for a given planning horizon. This thesis focuses on a scenario where a single-product type has to be delivered by a fleet of capacitated homogenous vehicles and housed at a depot over a finite and discrete planning horizon. The demand is fully available to the decision maker (supplier) at the beginning of the planning horizon, stock-outs are not allowed, and transportation costs and inventory holding costs of customers are taken into account in the objective function. Due to the NP-hard nature of the IRP, it is very difficult to develop an exact algorithm that can solve large-scale problems within a reasonable computation time. As an alternative, a hybrid evolutionary optimization algorithm based on two well-known meta-heuristics, the Genetic Algorithm and the Simulated Annealing Algorithm, is presented to handle the IRP. Namely, the Genetic Algorithm is related to the planning phase, while the Simulated Annealing Algorithm is associated with the routing phase. A repetitive procedure, containing characteristics from both referred meta-heuristics, is applied to obtain a near-optimal feasible solution. Testing instances with different properties are established to investigate algorithmic performance, and the computational results are then reported. Finally, a two-phase solution algorithm is presented to handle an extension of the IRP, the Inventory Routing Problem with Time Windows (IRPTW). The IRPTW, which has not been excessively researched in the literature, is a generalization of the standard IRP involving the added complexity that every customer should be served within a given time window. Α single-product type has to be delivered by a fleet of capacitated homogenous vehicles and housed at a depot over a finite and discrete planning horizon. The demand is fully available to the decision maker (supplier) at the beginning of the planning horizon, stock-outs are not allowed, and only transportation costs are taken into account in the objective function. The proposed two-phase solution algorithm is based on (a) a simple simulation for the planning phase and (b) the Variable Neighborhood Search Algorithm (VNS) for the routing phase. The computational study underscores the importance of integrating the inventory and vehicle routing decisions. Analytical results and graphic presentation formats are provided to convey meaningful insights into the problem.
Degree
thesis:*- Grantor dc:publisher
- Athens University Economics and Business (AUEB)
- Year dc:date
- 2017
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Lappas, Pantelis
- Λάππας, Παντελής
Subjects
dc:subject × 20- Δρομολόγηση
- Πρόβλημα Δρομολόγησης Αποθεμάτων
- Πρόβλημα Δρομολόγησης Αποθεμάτων με Χρονικά Παράθυρα
- Γενετικός αλγόριθμος
- Αλγόριθμος προσομοιωμένης ανόπτησης
- Εξελικτική Βελτιστοποίηση
- Προσομοίωση
- Αλγόριθμος Μεταβλητής Γειτονιάς Αναζήτησης
- Routing
- Inventory Routing Problem
- Inventory Routing Problem with Time Windows
- Genetic algorithm
- Simulated annealing algorithm
- Evolutionary Optimization
- Simulation
- Variable neighborhood search
- Κοινωνικές Επιστήμες
- Οικονομικά και Επιχειρήσεις
- Social Sciences
- Economics and Business
Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/40978