Bosch on the road

Fuel cells for mobility: HyPerformance – component kit for mobile fuel cell systems, Dr Helerson Kemmer

Organizer of Special Initiative on Autonomous System Design (ASD)

1) Microfluidics simulations in PEM fuel cells
2) Automated driving with infrastructure support
3) (Partly) automated risk assessment and security approval of versatile production machines
4) Adaptive machine intelligence for robot assembly

Talk about knowledge-infused learning for autonomous driving by Cory Henson

Experimental investigation on acoustics and efficiency of rotor configurations for electric aerial vehicles

Aerial vehicles based on distributed electric propulsion systems have gained great interest. Their rotors, however, create loud and annoying sound, which impends market success. Variations in rotor configuration can be observed on emerging concepts, whereby the main varied parameters are blade size, number of blades and blade distribution.

The focus of this paper is to identify how these parameters can be chosen to optimize eciency and acoustics, including psychoacoustic metrics of single rotors while hovering. Results from experimental investigations done in a hover-test-bench are presented. Rectangular, symmetric blades are used. Experiments are done varying blade size (radius 61 mm to 126 mm), number of blades (2 to 8) and blade distribution (equal and unequal angles). Acoustic measurements are analyzed regarding microphone position, sound pressure level, spectral characteristics and psychoacoustic metrics.

Results show that variations in blade size, number of blades and blade distribution can improve efficiency and acoustics. Influence of these parameters on the acoustic signature at constant thrust is discussed. Conclusions for optimized rotor design at aerial vehicles are derived and supplemented by resulting boundary conditions like assembly space and weight.

Verification of electric axle simulation models via a coupled eAxle-in-the-loop approach

The rapid change from internal combustion engines to electric drives in the automotive industry requires efficient development processes. The early phase of product development is increasingly accompanied by modelling and simulation of the complete vehicle. Extensive tests and optimizations on test benches follow. In this work, both the simulative and the test-based methods are improved by coupling a complete vehicle model with a test bench for electric drivetrains. Such an eDrive-in-the-Loop system enables realistic tests on the drive train test bench through the integration of the complete vehicle and the environment models. The modelling of electric drive trains is also associated with numerous uncertainties due to complex physical relationships. Particularly, thermal effects can only be represented in a very simplified manner in a system simulation. Outsourcing the drive train to the test bench can therefore significantly improve the simulation results.

In contrast to conventional hardware-in-the-loop systems, an object-oriented simulation model is used in this work. This enables a high level of detail and good understanding of the model structure but cannot be carried out on a real-time computer. Therefore, the connection to the test bench is established via a co-simulation with an integrated real-time interface. This requires a middleware that coordinates the data exchange between the simulation and the test bench in real time. With this approach, hardly any changes need to be made in the vehicle model. The co-simulation can be run on a Windows notebook, which means that no additional hardware is required.

With adequate coupling settings in the co-simulation middleware, a stable, accurate and real-time-capable test environment can be assured. The simulation model of the test vehicle can be modified very efficiently, thus allowing a fast analysis of effects on the electric axle due to changes on the (simulated) vehicle system level.