Abstract
KEYWORDS – Automated Driving, E/E Architecture, Power distribution, Fault tolerance, Short protection
ABSTRACT –
The introduction of highly automated driving with new functions like the highway assist calls for new E/E Architecture solutions. In contrast to now, the driver will no longer be able to react in case of any failure in the E/E systems. The necessity of fault tolerant functions imposes changes in the requirements for the E/E Architecture. A lot of effort focuses on the fault tolerance in sensors and actuators while the power distribution system has been widely neglected up to now, although fault tolerant power distribution at least for a limited amount of time is absolutely mandatory.
To consider all the requirements of a state of the art car, the starting point of this study is the architecture of a high class vehicle. This defines the technical requirements regarding power distribution, ampacity and voltage stability. Different solutions for a fault tolerant architecture have been developed. These architecture proposals have been integrated into a MATLAB / Simulink model of the vehicle. Based on the results of the simulation, they have been assessed with respect to their suitability to fulfill the requirements for fault tolerant operation. Afterwards, a prototype of the power distribution system including the power distribution nodes and the wiring harness has been built. With this prototype the functionality has been tested in a laboratory setup and in a real car environment.
Suitability of this approach has been shown through all stages of the used validation process. Namely, the simulation has shown the necessary voltage stability even when applying worst case load scenarios after a failure in the power distribution system. The hardware concept to detect and isolate failures has been derived. This concept is useable in different fault tolerant E/E topologies. Electronics simulation has successfully been passed and results were proven by testing with hardware prototypes. A model of the fault tolerant topology has been setup. Failures have been injected into this model and stability of the concept has been shown. The tests have been executed in a real vehicle environment.
Furthermore the impact of the incorporation of such a concept into an E/E Architecture of a high class vehicle in terms of costs and weight has been assessed. A market penetration scenario is derived from the results.
Duplicating infrastructure and components for a fault tolerant architecture is no longer reasonable due to cost, weight and space requirements. For a general introduction of automated driving functions the presented concept is a better solution.
The fault tolerant architecture concept has been advanced from a theoretical solution to real life validation with prototypes that have proven their capability in a vehicle environment. Based on these results the fault tolerant E/E Architecture will be put in a real vehicle.