Systems engineering has emerged as an approach that enables the successful realization of individual complex systems. In the automotive industry, product variety introduces a complexity that renders systems engineering models difficult to manage. Models Based Systems Engineering increases the level of abstraction and of detail to which systems are documented, thus providing the perfect context for reuse. Our purpose was to adopt "Product Line Engineering" techniques and enable efficient reuse of systems engineering assets, while satisfying process needs in an automotive industry context. The challenge is twofold: (i) to understand the needs of a mass production industry for which the reuse is usually focused towards physical parts and (ii) to satisfy the needs for capturing variability in complement to systems engineering models, by drawing on software product lines techniques. The main proposal of the thesis is a metamodel that extends OVM models in order to adopt and adapt variability modeling to MBSE needs. Furthermore, several subjects were developed around this theme of variability modeling in order to improve and complete the systems engineering framework: a derivation approach based on viewpoints and a set of recommendation-based heuristics for the configuration process. In collaboration with the CEA-LIST, a tool prototype was developed, which enabled us to refine, test and validate the contributions, in the industrial context at Renault.
| Identifer | oai:union.ndltd.org:CCSD/oai:tel.archives-ouvertes.fr:tel-01011186 |
| Date | 02 June 2014 |
| Creators | Dumitrescu, Cosmin |
| Publisher | Université Panthéon-Sorbonne - Paris I |
| Source Sets | CCSD theses-EN-ligne, France |
| Language | English |
| Detected Language | English |
| Type | PhD thesis |
Page generated in 0.0017 seconds