The measurement data were used, among others, to identify and validate the vertical dynamics part of the AFM’s high-fidelity vehicle model (cf. In several hundred separate experiments with different excitations, more than 50 measured variables were recorded in each case. They enable recording of many important quantities. In addition to the test rig sensors, the AFM is equipped with its own vertical dynamics sensors as described in Section 3.3. Its sensor system allows the measurement of the post position, speed, accelerations and wheel loads. Moreover, it is possible to install additional sensors under unvarying conditions of the test rig. In this way, driving on precisely defined road profiles (such as, for instance, synthetic profiles according to ISO 8608 ) can be simulated in a reproducible manner. The posts induce a vertical movement of the vehicle. To address these issues, the AI-For-Mobility (AFM) project was started at the DLR Institute of System Dynamics and Control (SR).ĭuring this process, the vehicle is placed on four hydraulically driven vertical posts, each following its predefined position signals. Furthermore, such concept vehicles usually do not have road approval, so that the developed functions cannot be tested in challenging road traffic scenarios, but only on secured test sites. Although novel control methods can be well represented by the ROMO’s four separate wheel-robots, investigations can only be transferred to road vehicles to a limited extent. Prototypical vehicles such as the ROMO are fundamentally different from ordinary production vehicles in terms of their basic architecture and driving characteristics. Among others, the ROboMObil (ROMO), a robotic and over-actuated electric vehicle, which can control the steering angles and in-wheel torques by means of four-wheel robots independently, is available for testing. The Department of Vehicle System Dynamics at the German Aerospace Center (DLR) is researching novel control algorithms for vehicles. This paper presents the vehicle’s design and architecture in a detailed manner and shows a promising application potential of AFM in the context of AI-based control methods. This makes it an attractive research and test platform, both for automotive applications and technology demonstrations in other scientific fields (e.g., robotics, aviation, etc.). Despite all modifications, it is approved for public road use and meets the driving dynamics properties of a standard road vehicle. A full by-wire control system enables the vehicle to be used for applications in the field of automated driving. ![]() They measure the vehicle’s state holistically and perceive the surrounding environment, while high performance on-board CPUs and GPUs handle the sensor data. Since AI-based methods are data-driven, the vehicle is equipped with manifold sensors to provide the required data. A production hybrid vehicle serves as a base platform. AI-For-Mobility (AFM) is the new research platform to investigate and implement novel control methods based on Artificial Intelligence (AI) within the Department of Vehicle System Dynamics at the German Aerospace Center (DLR).
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |