Search Results

As part of a system model, a PEM fuel cell stack model is presented for functional tests and pre-calibration of control units on hardware-in-the-loop (HiL) test benches. From the basic idea to couple a 1-D membrane model with a spatially distributed abstraction of the gas channel, a real-time capable 1-D+1-D PEM FC stack model is constructed. Fundament for the HiL usage is an explicit formulation of the commonly implicit model equations. With that, not only calculation time can be reduced, but also model accuracy is preserved. A validation using test bench data emphasizes the accuracy of the model. Finally, a runtime and eigenvalue analysis of the stack model proves the real-time capability.

This report presents a new compensation method for distortions related to dead time, caused by B6-inverters with pulse-width-modulated output voltages. In spite of low sampling rates, the new method of compensation is effective at all ranges of rotation speed up to the field weakening region. No additional hardware is required for its implementation. The effectiveness of the new method has been shown experimentally. A description of the relevant distortions is given first to provide a basis for the development. This considers the field weakening region, and offers an illustrative method of quantifying the distortions. It is also shown that the use of compensation methods that do not take the sampling time into account leads to additional distortions. It is even possible that they exceed the distortions in an equivalent system without compensation.

Modern automotive development evolves beyond artificial intelligence for highly automated driving, and toward an interconnected manifold of data-driven development processes. Widely used analytical system modelling struggles with rising system complexity, invoking approaches through data-driven system models. We consider these as key enablers for further improvements in accuracy and development efficiency. However, literature and industry have yet to thoroughly discuss the relevance and methods along the vehicle development cycle. We emphasize the importance of data-driven system models in their distinct types and applications along the developing process, from pre-development to fleet operation. Data-driven models have proven in other works to be fast approximators, of high accuracy and adaptive, in contrast to physics-based analytical approaches across domains.