Advertisement

A methodology for the optimized design of an E/E architecture component platform

  • Sebastian Graf
  • Michael Glaß
  • Jürgen Teich
  • Daniel Platte
Conference paper
Part of the Proceedings book series (PROCEE)

Abstract

In recent years, more and more cars are built upon mechanical platforms and, thus, enable to reduce development time and cost due to increased reuse of already developed mechanical components like combustion engines or suspensions. But, to support a seamless platform-based development of the whole car, also the Electric/Electronic (E/E) architecture and the included vehicle electronics have to be build based on E/E architecture component platforms. There, component manifestations like different hardware versions of an Electronic Control Unit (ECU) should be reused whenever possible and beneficial. To enable this, the work at hand proposes to use a multi-variant Design Space Exploration that enables an automatic multi-objective optimization of the E/E architecture component platform as well as the individual variants’ E/E architectures at once. This allows to enhance the E/E architecture design process by means of design automation approaches. Based on previous work proposing a symbolic encoding of the overall optimization problem, practically relevant design space restrictions are proposed and efficiently integrated in an existing multi-variant problem encoding. In combination with the multi-objective optimization, this enables the developer to ponder between several important design objectives while automatically ensuring the validity of the obtained solutions with respect to real-world design restrictions. Within the context of an INI.FAU project, the proposed methodology is used to model real-world use cases from the automotive safety domain and enables to automatically optimize upcoming vehicle safety-architecture component platforms for future cars.

Keywords

Design Space Electronic Control Unit Design Space Exploration Reduce Development Time Symbolic Encode 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Copyright information

© Springer Fachmedien Wiesbaden 2015

Authors and Affiliations

  • Sebastian Graf
    • 1
  • Michael Glaß
    • 1
  • Jürgen Teich
    • 1
  • Daniel Platte
    • 2
  1. 1.University of Erlangen-NürnbergErlangen-NürnbergGermany
  2. 2.AUDI AGIngolstadtGermany

Personalised recommendations