Skip to Main content Skip to Navigation
Theses

Développement d’une nouvelle famille d’indicateurs de performance pour la conception d’un système manufacturier reconfigurable (RMS) : approches évolutionnaires multicritères

Abstract : The modern manufacturing environment is facing a paradigm shift that require more changeability at physical and logical levels. A Changeable Manufacturing System is defined as a production system that has the ability to facilitate the right changes, allowing the adjustment of its structures and processes in response to the different needs. In this context, manufacturing systems must have a very high level of reconfigurability, which is considered to be one of the major enablers of changeability. From the perspective of the “Factory of the future”, the reconfigurability is essential to effectively adapt to the ever-increasing complexity of manufacturing environments. It allows a rapid, efficient and easy adaptation of these systems while being responsive, robust and economically competitive. The objective is to respond to new internal and external constraints in terms of globalization, variety of products, mass customization, and shorter lead times. Through this thesis, we study the problem of design of reconfigurable manufacturing systems (RMS) that meets these requirements. The goal is to design responsive systems based on their key features of reconfigurability. We have studied the RMS design problem on three levels: (i) the level of the components, relating to the modules of the reconfigurable machines, (ii) the machine level and their interactions, as well as the impact of these interactions on the system and (iii) the workshop level composed of all the reconfigurable machines. We have developed for each level, performance indicators to ensure a better responsiveness and a high performance of the designed system, like the modularity index, the flexibility index, the robustness index and the layout evolution effort of a reconfigurable system. For each of the studied problems, we developed multicriteria optimization models, solved through heuristics or multicriteria metaheuristics (such as archived multi-objective simulated annealing (AMOSA) and multi-objective genetic algorithms (NSGA-II)). Numerous numerical experiments and analyzes have been performed to demonstrate the applicability of our approaches
Complete list of metadatas

https://tel.archives-ouvertes.fr/tel-01886887
Contributor : Abes Star :  Contact
Submitted on : Wednesday, October 3, 2018 - 12:39:07 PM
Last modification on : Friday, May 17, 2019 - 11:39:45 AM
Long-term archiving on: : Friday, January 4, 2019 - 1:50:14 PM

File

DDOC_T_2018_0112_HADDOU_BENDER...
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-01886887, version 1

Collections

Citation

Hichem Haddou Benderbal. Développement d’une nouvelle famille d’indicateurs de performance pour la conception d’un système manufacturier reconfigurable (RMS) : approches évolutionnaires multicritères. Systèmes et contrôle [cs.SY]. Université de Lorraine, 2018. Français. ⟨NNT : 2018LORR0112⟩. ⟨tel-01886887⟩

Share

Metrics

Record views

211

Files downloads

211