Model-based optimization of automotive E/E-architectures
Stefan Kugele, Gheorghe Pucea · 2014
In this paper we present a generic framework to enable constraint-based automotive E/E-architecture optimization using a domain-specific language. The quality of today's automotive E/E-architectures is highly influenced by the mapping of software to executing hardware components: the so-called deployment problem. First, we introduce a holistic architectural model facilitating a seamless model-based development from requirements management to deployment, which is the focus of this work. Second, we introduce our domain-specific constraint and optimization language AAOL (Automotive Architecture Optimization Language) capable to express a wide range of deployment-relevant problems. Third, we present a generic, i.e., solver-independent framework currently supporting multi-objective evolutionary algorithms (MOEA). We investigate the feasibility of the approach by dint of a case study taken from the literature.