Browse Publications Technical Papers 2017-26-0358

FMI for Physics-Based Models on AUTOSAR Platforms 2017-26-0358

As automobiles become increasingly smarter, the need to understand within the automotive software the physical behavior of its parts is growing as well. The laws of physics governing such behavior are mostly formulated as differential equations, which today are usually created or obtained from various modeling tools. For solving them, the tools offer several solvers to satisfy the requirements of different problems. E.g. simple and fast explicit low order solvers for non-stiff problems and more complex implicit solvers for stiff problems.
Though the modeling and code generation features as available in such tools are desirable for embedded automotive software, they cannot be used directly due to special restrictions with respect to hard realtime constraints. One such restriction is the organization of automotive software in components complying with the AUTOSAR standard which is not widely supported by the modeling tools. More importantly, AUTOSAR does not provide any solvers or a solver interface to be used by models. In the offline world, this is offered by the Functional Mock-up Interface (FMI) which is becoming increasingly popular and is supported by most of the relevant tools.
In this paper, we propose an interface between FMI and AUTOSAR, offering the combination of models with solver libraries. With this, source code could be generated from various established modeling tools and directly be imported into existing automotive software projects. Furthermore, the definition of this interface allows the creation and use of solver libraries on the embedded platform, enabling online simulations even of numerically challenging physical processes. Results from a prototypical implementation will be presented.


Subscribers can view annotate, and download all of SAE's content. Learn More »


Members save up to 18% off list price.
Login to see discount.
Special Offer: Download multiple Technical Papers each year? TechSelect is a cost-effective subscription option to select and download 12-100 full-text Technical Papers per year. Find more information here.