The lowest layer of the scalable vehicle dynamics control architecture is assigned to the actuators. It comprises the physical execution of the control commands from the overlying motion execution layer. The institute’s research activities has been oriented to real-time modeling and code generation by means of the Modelica and Functional Mockup Interface (FMI) technologies within several cooperation research projects together with industrial partners (mainly Robert Bosch GmbH). The focus has been on enabling and testing a tool chain for model-based controller development for automotive embedded targets. Together with Dassault Systèmes, new technologies have been developed to solve nonlinear inverse dynamic models and nonlinear feedback linearization controllers under hard real-time requirements. Within the international ITEA project EMPHYSIS that was initiated by Bosch and the institute, the institute and other partners are currently developing a novel embedded FMI (eFMI) standard.
This interface standard will facilitate the exchange of physics-based nonlinear models between modeling and simulation environments with software development environments for ECU and other embedded systems. In a use case of the project, the institute has prepared a tool chain to implement both a state estimator and a controller for a high-performance semi-active suspension damper on a predevelopment ECU. The tool chain, as depicted in figue below, is based on controller and prediction models in Modelica, generated eFMI code by Dymola (Algorithm code) and TargetLink (C-code), and an institute’s library of C-functions dedicated to implementation of nonlinear Kalman filter variants on embedded systems. All the C-code parts are configured in an AUTOSAR environment and finally compiled for the target ECU.