National Technical University of Athens (NTUA)
Μοντελοποίηση και επίλυση σύνθετων προβλημάτων προγραμματισμού εργασιών με χρήση χρονισμένων αυτομάτων: εφαρμογή σε δυναμικά δίκτυα παραγωγής
Abstract
dc:descriptionThe thesis aims at developing an integrated innovative approach for solvingproduction scheduling problems by exploiting the computational capabilities of finiteautomata and more specifically of a new extension of them, called timed automata.The characteristics and capabilities of timed automata in conjunction with thedevelopment of tools which allow the accurate modelling of complex systems basedon the principles of the automata theory were the main reasons for examining, inthe framework of the thesis, whether such a modern approach could be appliedeffectively on problems related to job and task scheduling.In the above context, the thesis proposes an integrated approach for modellingproduction scheduling problems based on the visualization of the several discretestates of a production system. Particular emphasis is given to the development ofmodels of production systems with complex structures, which significantly differfrom the usual production systems’ structures on which the majority of schedulingapproaches are focused.More specifically, in the context of the thesis, a comprehensive study is presentedbased on existing applications of timed automata on production schedulingproblems, while a new innovative approach is being proposed for modelling complexproduction systems in order to solve scheduling problems by utilising automatatheory optimisation algorithms. A major innovation of the thesis is the addition ofseveral types of costs to the timed automata models developed for solving jobscheduling problems. This way it is possible to address and optimally solve problemsthat the objective function includes not only time but also cost elements involved inthe production process, such as processing costs, transportation costs, storage costs,production switching costs etc.In the above framework, several timed automata models have been developedwhich can be used as templates for building automata networks able to deal withcomplex cases like preemptive job scheduling problems and scheduling of jobs withalternative production routes. The optimization criteria that can be used to solveproblems using the proposed approach are not limited to the usual time-relatedcriteria but they involve a variety of cost elements, providing the possibility offinding optimal schedules that minimize either total production cost or specific typesof costs, depending on the case or the problem.The proposed approach is extended and adapted appropriately to addressscheduling problems in Dynamic Manufacturing Networks (DMN) environments. Theinnovative approach, presented in the thesis, addresses the problem of the limited coordination between producers belonging in a manufacturing network, taking intoaccount several types of production and logistics costs, in order to optimize theoperation and the production plans of the whole network.The proposed approach has been applied successfully, in the context of the thesis,on a real manufacturing network case. More specifically, the manufacturing networkexamined is a network formed by a Greek Original Equipment Manufacturer (OEM),producing heating boilers, and by its main suppliers. The timed automata approachhas been applied on real scheduling problems of this network and the results werevery promising since the production plans produced were significantly better thanthe plans which were initially formed by the OEM.By examining the approach proposed in the thesis, and by taking into account theexisting trend of utilising automata networks in manufacturing environments forsolving several types of problems, it can be concluded that the thesis has significantvalue not only in the research field but also in practice. The continuous developmentof automata modelling tools in conjunction with the fact that the proposed approachprovides an easy-to-apply method for modelling complex production systems andsolving scheduling problems guarantee that the outcomes of the thesis can befurther exploited both in research and commercial level.
Degree
thesis:*- Grantor dc:publisher
- National Technical University of Athens (NTUA)
- Year dc:date
- 2014
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Panopoulos, Dimitrios
- Πανόπουλος, Δημήτριος
Subjects
dc:subject × 16- Χρονοπρογραμματισμός παραγωγής
- Χρονισμένα αυτόματα
- Ανάθεση/δρομολόγηση εργασιών
- Μοντελοποίηση παραγωγικών συστημάτων
- Δυναμικά δίκτυα παραγωγής
- Production scheduling
- Timed automata
- Production systems modelling
- Job assignment
- Dynamic manufacturing networks
- Επιστήμες Μηχανικού και Τεχνολογία
- Επιστήμη Ηλεκτρολόγου Μηχανικού, Ηλεκτρονικού Μηχανικού, Μηχανικού Η/Υ
- Επιστήμη Μηχανολόγου Μηχανικού
- Engineering and Technology
- Electrical Engineering, Electronic Engineering, Information Engineering
- Mechanical Engineering
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/43683