Modelling and schedulling of remanufacturing systems

In this thesis, the modelling and scheduling of remanufacturing systems is considered. An eXtended two-level Coloured Petri Net (XCPN) model is proposed for the objective of modelling remanufacturing shops. The extended model integrates the formalism of an existing two-level Coloured Petri net model and the object-oriented modelling approach, which includes the concept of modules. This modular concept is different from existing ones in that the model development is easy and the developed models are flexible and highly reusable. In order to provide the modeller with a systematic procedure for model development, a modelling methodology based on XCPN is proposed. The proposed methodology ensures clear separation of resource modelling from the modelling of process plans and parts. A software prototype is developed for building XCPN models and for their quantitative analysis by simulation. Remanufacturing is considered as a production with high level of uncertainties (uncertainties in the quantity and quality of used products and uncertainties in process plans and processing times). Modelling approaches for highly uncertain environments are analysed and a completely reactive approach based on Priority Dispatching Rules (PDR) is selected. Three new PDRs are proposed with the goal to minimise tardiness related performance measures. Extensive simulation experiments are conducted in order to evaluate the performance of the new and existing rules in several environments, which are typical of remanufacturing shops. In many cases, the new rules outperform existing rules or are not significantly worse than the best of them with respect to the tardiness related performance measures. At the same time, existing PDRs are shown to be applicable in the remanufacturing environment especially with respect to the flow-time related performance measures. In order to show the effectiveness of the new rules they are tested in two case studies of real-life remanufacturing systems using the proposed modelling methodology and software.

Xirouchakis, Paul
Lausanne, EPFL
Other identifiers:
urn: urn:nbn:ch:bel-epfl-thesis3343-0

Note: The status of this file is: EPFL only

 Record created 2005-08-31, last modified 2018-05-01

Texte intégral / Full text:
Download fulltext

Rate this document:

Rate this document:
(Not yet reviewed)