Datenbestand vom 13. März 2019
Tel: 089 / 66060798
Mo - Fr, 9 - 12 Uhr
Fax: 089 / 66060799
aktualisiert am 13. März 2019
978-3-8439-2356-9, Reihe Informatik
Design and Optimization of Multi-Variant Automotive E/E Architecture Component Platforms
195 Seiten, Dissertation Universität Erlangen-Nürnberg (2015), Softcover, A5
The number of car models and variants offered to customers is steadily increasing. To reduce development costs, car manufacturer more and more tend to implement several new car variants based on one common platform. Initially started from a mechanical platform-based design, it is even more required to also cover the electric and electronic (E/E) architecture, especially regarding vehicle electronics. The design of one E/E architecture already is a challenging task. The complexity of the design phase even more increases in case of covering not only one single car, but a huge variety of car variants build upon one common E/E architecture component platform.
To deal with platform-based design of vehicle electronics, this thesis proposes a holistic approach to apply electronic design automation and design space exploration to determine an optimized E/E architecture component platform covering vehicle electronics of several car variants by applying a multi-variant optimization. Beside major contributions in the area of symbolic model encoding as the basic requirement for applying enhanced optimization techniques in the automotive domain, a powerful modeling approach for specifying the overall component platform optimization problem itself is proposed. By means of separation of a variant into functional and architectural point of view, an implicit functional variants model enables a compact expression of the overall variety of possible car configurations. By selecting a set of functional variants, each representing one car variant, a design space exploration is applied to determine recommendations for the design of the E/E architecture component platform, e.g., wrt. number and individual configuration of manifestations of the involved ECUs as well as their allocation to variants. As design automation and optimization requires to rate each observed solution, this thesis proposes new design objectives and also more complex rating like robustness of an objective against uncertain parameters or simulation-based timing analysis.
Finally, as parts of this work were performed within an INI.FAU collaboration with the department of safety electronics at AUDI AG in Ingolstadt, real-world case studies give evidence of the proposed approaches.
The results highlight the efficiency and applicability of the proposed approaches for the design of upcoming E/E architecture component platforms as a key enabler for the automatic optimization of architecture component platforms.