University of Patras
Design and implementation of mathematical metaheuristic algorithms for solving the mathematical transportation problem in deterministic and fuzzy environments
Abstract
dc:descriptionThis doctoral thesis elucidates the research conducted in addressing the Transport Problem, a crucial component in the functioning of industrial enterprises as well as in both public and private transportation sectors. To address the problem, two novel algorithms rooted in the Particle Swarm Optimization (PSO) framework were devised and executed. The first proposed algorithm, TrigAC-PSO, empowered each individual particle, instilling a sense of autonomy and confidence, enabling them to independently navigate toward optimal positions while leveraging their own experiences for enhanced learning. Beginning with a modest influence from its cognitive component while allowing the social component to dominate, the particle gradually gains greater autonomy as the iterations progress, breaking free from the constraints imposed by the swarm. By the final iteration, the particle consolidates its experiences from the optimal positions previously encountered and re-aligns itself with the swarm, engaging in a reciprocal exchange of information with the other members. This culminates in a harmonious equilibrium of knowledge and movement. It is significant to highlight that this research introduces a pioneering method that empowers each particle to leverage its unique capabilities and strengths, allowing it to navigate the solution space independently of the swarm's guidance. The second proposed variation, FSVAC, exhibits a high degree of combinatorial complexity, as it incorporates contributions from both individual particles and the swarm at regular intervals throughout the iterations. These two variations were rigorously evaluated across 32 different problems and subsequently compared with established algorithms, where they demonstrated their superiority by delivering solutions distinguished by absolute completeness and precision. The results obtained from the specific algorithms represent the most impressive outcomes to date for the particular problems examined, with TricAC-PSO rightfully claiming the top position. The research focusing on the effectiveness of TricAC-PSO, we conducted experiments under conditions of uncertainty, modeling transportation costs and problem solutions using fuzzy numbers. TricAC-PSO was tested across 28 scenarios involving both classical and generalized fuzzy numbers. Subsequently, it was benchmarked against both well-established and contemporary algorithms, yielding highly satisfactory results that provided near-optimal solutions to intricate and complex challenges. Ultimately, the esteemed Urban Transit Routing Problem was a significant focus of this doctoral thesis. The application of Particle Swarm Optimization algorithm to well-documented road networks in the international literature yielded remarkable findings regarding multi-route challenges. This innovative approach successfully met passenger expectations while minimizing travel time, delivering an impressive solution to complex routing issues.
Degree
thesis:*- Grantor dc:publisher
- University of Patras
- Year dc:date
- 2024
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Aroniadi, Chrysanthi
- Αρωνιάδη, Χρυσάνθη
Subjects
dc:subject × 21- Πρόβλημα μεταφοράς
- Αλγόριθμος βελτιστοποίησης σμήνους σωματιδίων
- Πρόβλημα δρομολόγησης αποθεμάτων
- Ασαφής λογική
- Transportation problem
- Particle swarm optimization
- Fuzzy logic
- Fuzzy logic and fuzzy system
- Transportation networks
- Μαθηματικά
- Φυσικές Επιστήμες
- Εφαρμοσμένα μαθηματικά
- Επιστήμη Ηλεκτρονικών Υπολογιστών και Πληροφορική
- Πληροφοριακά συστήματα
- Τεχνητή νοημοσύνη
- Mathematics
- Natural Sciences
- Applied Mathematics
- Computer and Information Sciences
- Information Systems
- Artificial Intelligence
Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/58318