OEM&Lieferant Ausgabe 2/2020

84 A project report on the AVL DRIVINGCUBE ™ From vision to reality A highly dynamic vehicle-in-the-loop testbed in use at the Technical University of Berlin By Dr. Christian Schyr, Project Manager Principal Engineer, AVL Deutschland GmbH The Automotive Engineering Department at the Technical University of Berlin headed by Prof. Dr.-Ing. SteffenMüller planned to purchase a highly dynamic vehicle-in-the-loop testbed (ViL testbed) in 2016 for the examination, analysis, research and development of vehicle guidance control, powertrain and chassis control systems as well as energy management systems of vehicles. The testbed nowadays represents the heart of the Automotive Engineering Test Center at the TU Berlin on the Wedding Campus. The testbed concept is based on a 4WD pow- ertrain testbed with a vehicle as a whole. Highly dynamic synchronous machines are used to ensure real operating conditions – up to the limits of driving dynamics. Steering in- terventions are made possible by the mechani- cal decoupling of the steering rod on thewheel carrier and the attachment of a linear steering actuator. The test vehicle is connected to the virtual environment via sensor interfaces. This ensures that the control units in the vehicle are suppliedwith all necessary environmental and driving dynamics parameters. A battery simu- lator is used for the efficient development and evaluation of innovative operating strategies on hybrid and electric vehicles. Commissioning took place in spring 2018, immediately followed by the opportunity to use the testbed in the RobustSENSE research project in cooperation with the Fraunhofer Institute for Open Communication Systems (FOKUS) in Berlin. The task was the inte- gration and validation of a new platform for robust and reliable environment detection for driver assistance systems and highly au- tomated driving even under adverse weather conditions. For functional verification and demonstration of the project results, the engineers equipped a research vehicle with a robust system archi- tecture for the two driving functions Adap- tive Cruise Control (ACC) and Lane Keep Assist (LKA). In addition to the standard sensors, further reference sensors (Lidar and Camera) were installed for environment detection. In real driving tests under winter conditions on country roads and motorways, the scenarios critical for environment recognition and driv- ing dynamicswere identified and then virtually mapped for the AVL DRIVINGCUBE™. Accord- ing to the validation method of RobustSENSE, the project participants replaced the real vehi- cle sensors on the testbedwith sensor models. Images: © AVL Deutschland GmbH Innovations In this case, sensor models were implemented for the detection of dynamic objects and road lines as well as for data from high-resolution maps, localization via satellite navigation and Car2X communication. The behavior of the automated functions could be tested on the testbed in various weather and traffic conditions under real loads on the drive, brake and steering systems as well as the associated control units and displays in the cockpit. Due to the simultaneous visualization of the recorded environment and the detected objects in the RobustSENSE platform, their function and reliability could be tested very efficiently on the testbed. With the vehicle-in-the-loop testbed AVL DRIVINGCUBE™, the TU Berlin is now able to carry out comprehensive research and tests on self-propelled and electrified vehicles. The new technology makes it possible to investi- gate not only classic propulsion systems but also the propulsion and energy management of e-vehicles and autonomous driving func- tions. The implemented concept of a vehicle testbed for the operation of both the drivetrain as well as the steering and braking system in combi- nation with the simulation of environment