Πλοήγηση ανά Επιβλέπων "Mourtos, Yiannis"

A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

Α Β Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ Ν Ξ Ο Π Ρ Σ Τ Υ Φ Χ Ψ Ω

Τώρα δείχνει 1 - 2 από 2

Bus driver scheduling problem - A constraint programming model developed with python
(2022) Aznavouridis, Konstantinos; Αζναβουρίδης, Κωνσταντίνος; Androutsopoulos, Konstantinos; Zachariadis, Emmanouil; Mourtos, Yiannis
Bus driver crew scheduling poses a great challenge for bus operators around the world. Solving the problem efficiently and optimally is the subject of many studies of operations research. Lately, constraint programming has shown promising results in this field due to the fact that it possesses special characteristics for this kind of scheduling problems and it uses more programming-friendly expressions. The increase of computing power further facilitates constraint programming because this kind of models, and especially those regarding scheduling problems, can have thousands of constraints. The aim of this thesis is to develop a python tool which incorporates constraint programming modelling methodologies to solve the single-depot and multiple-depot bus driver crew scheduling problem. Lower and upper bounds, the later with greedy algorithm, were calculated in order to reduce the time complexity of our solutions. Real world data instances were used to test the developed models. Finally, python was chosen because it is an easy language for rapid prototyping and development with the ability to form complete multi-platform solutions easily.
Integrated methods and systems for optimization and decision support
(30-09-2017) Plitsos, Stathis; Athens University of Economics and Business, Department of Management Science and Technology; Magos, Dimitrios; Tarantilis, Christos D.; Doukidis, George I.; Εμίρης, Δημήτριος; Δούμπος, Μιιχαήλ; Γιαννίκος, Ιωάννης; Mourtos, Yiannis
First we focus on the multi-index assignment. We propose several components that can be employed across different types of assignment, i.e, a constraint propagation mechanism, a tabu-search meta-heuristic, a new variant of the Feasibility Pump heuristic that employs cutting planes, along with a new Branch & Cut method. Results show that these components when employed together reduce the time to optimality or the integrality gap for large instances compared to a commercial solver. Furthermore, this work paves the way towards the development of a DSS, which can facilitate several types of use. The second problem is the energy-aware production scheduling. Here, we present an energy-aware production scheduling DSS as designed, implemented and evaluated in a real context. In short, this work contributes to decision support for energy-efficient manufacturing by a metaheuristic algorithm that hierarchically optimizes flexible job-shop scheduling problems, a set of data requirements and the DSS evaluation in real settings. Last, we focus on the the binary multi-dimensional knapsack problem. Here, we describe a new primal-dual method. Current exact approaches and commercial solvers run into difficulties even for a small-to-medium number of constraints and variables. The proposed primal-dual method employs the linear relaxation, enhanced by global lifted cover inequalities to improve the upper bound and a new version of the Feasibility Pump heuristic that uses these cuts in the pumping procedure to obtain better and feasible lower bounds.