Institut für Innovative Mobilität (IIMo)
Refine
Year of publication
Document Type
- Conference Paper (129)
- Article (71)
- Part of a Compilation / Book Chapter (1)
- Preprint (1)
Institute
- Institut für Innovative Mobilität (IIMo) (202)
- Fakultät Elektro- und Informationstechnik (122)
- Fakultät Elektrotechnik und Informatik (58)
- Fraunhofer-Anwendungszentrum "Vernetzte Mobilität und Infrastruktur" (22)
- AImotion Bavaria (4)
- CARISSMA Institute of Automated Driving (C-IAD) (1)
- CARISSMA Institute of Electric, Connected and Secure Mobility (C-ECOS) (1)
- Fakultät Informatik (1)
- Fakultät Maschinenbau (1)
Review
- peer-review (137)
- ja (46)
- nein (10)
- editorial review (7)
Version
- published (102)
Enhancing traffic signal optimisation has the potential to improve urban traffic flow without the need for expensive infrastructure modifications. While reinforcement learning (RL) techniques have demonstrated their effectiveness in simulations, their real-world implementation is still a challenge. Real-world systems need to be developed that guarantee a deployable action definition for real traffic systems while prioritising safety constraints and robust policies. This paper introduces a method to overcome this challenge by introducing a novel action definition that optimises parameter-level control programmes designed by traffic engineers. The complete proposed framework consists of a traffic situation estimation, a feature extractor, and a system that enables training on estimates of real-world traffic situations. Further multimodal optimisation, scalability, and continuous training after deployment could be achieved. The first simulative tests using this action definition show an average improvement of more than 20% in traffic flow compared to the baseline – the corresponding pre-optimised real-world control.
It is well known that lithium-ion batteries pose a certain safety risk. The thermal runaway of a cell and the subsequent thermal propagation through the battery are considered particularly dangerous. Effective solutions for their mitigation are therefore of great interest. Previous studies have shown the significant influence
of a cell’s state of charge (SOC) on its behavior during thermal runaway. This relation may be exploitable in a battery pack to improve its safety. This study aims to assess the leverage of active SOC reduction in the imminent threat of thermal runaway. Implementing such a technique could become feasible with the emergence of reconfigurable battery systems. Four experiments were conducted, each with a module of three fresh 63 Ah high energy pouch cells in a spring-loaded bracing. The experiments studied different stationary SOC configurations, uniform (100% and 60%) and non-uniform (100%–60%–100% and 100%–20%–100%). The results indicate that thermal propagation is substantially delay (87 s) by discharging a cell in its path. The SOC reduction primarily decreases the maximum temperature of the respective cell. Further effects are a calmer thermal runaway and prolonged propagation time within the cell as well as to the next cell. In comparison, the SOC reduction has little impact on the cell’s own triggering time, as the triggering time is mainly determined by the thermal energy transferred from the preceding cell and hence by its SOC. Furthermore, the analysis of the experimental data (temperature, voltage, pressure, video) gives insights into the propagation of thermal runaway through the individual layers of a cell. With reference to the position of a cell relative to the origin of the thermal propagation, a decrease of its mass loss and an increase of its internal propagation time is observed. This effect is attributed to the decreasing module pressure due to progressive loss of material. The assessment shows that active SOC reduction techniques have great leverage for mitigating or even stopping thermal propagation in a battery pack.
Environment perception using camera, radar, and/or lidar sensors has significantly improved in the last few years because of deep learning-based methods. However, a large group of these methods fall into the category of supervised learning, which requires a considerable amount of annotated data. Due to uncertainties in multi-sensor data, automating the data labeling process is extremely challenging; hence,
it is performed manually to a large extent. Even though full automation of such a process is difficult, semiautomation can be a significant step to ease this process. However, the available work in this regard is still very limited; hence, in this paper, a novel semi-automatic annotation methodology is developed for labeling RGB camera images and 3D automotive radar point cloud data using a smart infrastructure-based sensor setup. This paper also describes a new method for 3D radar background subtraction to remove clutter and a new object category, GROUP, for radar-based object detection for closely located vulnerable road users. To validate the work, a dataset named INFRA-3DRC is created using this methodology, where 75% of the
labels are automatically generated. In addition, a radar cluster classifier and an image classifier are developed, trained, and tested on this dataset, achieving accuracy of 98.26% and 94.86%, respectively. The dataset and Python scripts are available at https://fraunhoferivi.github.io/INFRA-3DRC-Dataset/.
Collaborative perception in automated vehicles leverages the exchange of information between agents, aiming to elevate perception results. Previous camera-based collaborative 3D perception methods typically employ 3D bounding boxes or bird's eye views as representations of the environment. However, these approaches fall short in offering a comprehensive 3D environmental prediction. To bridge this gap, we introduce the first method for collaborative 3D semantic occupancy prediction. Particularly, it improves local 3D semantic occupancy predictions by hybrid fusion of (i) semantic and occupancy task features, and (ii) compressed orthogonal attention features shared between vehicles. Additionally, due to the lack of a collaborative perception dataset designed for semantic occupancy prediction, we augment a current collaborative perception dataset to include 3D collaborative semantic occupancy labels for a more robust evaluation. The experimental findings highlight that: (i) our collaborative semantic occupancy predictions excel above the results from single vehicles by over 30%, and (ii) models anchored on semantic occupancy outpace state-of-the-art collaborative 3D detection techniques in subsequent perception applications, showcasing enhanced accuracy and enriched semantic-awareness in road environments.
Fault testing in the production line of automotive traction machines is essential to ensure the desired lifetime. Since repetitive partial discharges (PDs) caused by anomalies in the insulation system lead to premature breakdowns of electrical machines, a reliable PD detection is of great importance. This paper proposes deep learning (DL) methods to improve the discrimination of PD from background noise in comparison with the state-of-the-art amplitude based PD detection in the production line. First, a systematic data extraction and labeling procedure is introduced to obtain correctly labeled datasets from arbitrary PD measurements. In addition, datasets are enhanced with low signal-to-noise ratio PD pulses by applying a special data augmentation approach. 13 different convolutional, recurrent and fully connected neural networks are compared for various time-frequency representations of the input signals. Hyperparameters for input transform, network topology and solver are optimized for all 13 combinations to ensure a fair case study. As a result, the two-dimensional convolutional neural network with continuous wavelet transform achieves the best accuracy of around 99.76% on a test dataset of PD signals originating from previously not utilized test objects. All DL models considered in this comparison outperform the state-of-the-art threshold-based PD classification. Even for PD events with an amplitude close to the noise level, the detection rate is still around 95% for the best network. Furthermore, without applying the proposed data augmentation procedure, the DL models investigated are not able to distinguish small PD pulses from noise.
Evaluating the Aging-Induced Voltage Slippery as Cause for Float Currents of Lithium-ion Cells
(2023)
This paper provides a comprehensive exploration of float current analysis in lithium-ion batteries, a promising new testing method to assess calendar aging. Float currents are defined as the steady-state trickle charge current after a transient part. In the literature, a correlation to capacity loss was reported. Assuming the float current compensates for the voltage decay over time and is linked to calendar aging, effects from voltage slippery must be considered. The dU/dQ analysis suggests solely a loss of active lithium. Therefore, we investigate the solid electrolyte interphase (SEI) growth as the general aging mechanism to explain the origin of float currents. Our results show that the voltage slippery theory holds true within the low to middle test voltage ranges. However, the theory’s explanatory power begins to diminish at higher voltage ranges, suggesting the existence of additional, yet unidentified, factors influencing the float current. A shuttle reaction or lithiation of the cathode by electrolyte decomposition are the most promising alternative aging mechanisms at high voltages. The paper proposes a unique voltage slippery model to check for correlations between aging mechanisms, the float current test and the check-up test. For a better understanding, test strategies are proposed to verify/falsify the aging mechanisms beyond SEI.
Cameras mounted on intelligent roadside infrastructure units and vehicles can detect humans on the road using state-of-the-art perception algorithms, but these algorithms are presently not trained to distinguish between human and adult. However, this is a crucial requirement from a safety perspective because a child may not follow all the traffic rules, particularly while crossing the road. Moreover, a child may stop or may start playing on the road. In such situations, the separation of a child from an adult is necessary. The work in this paper targets to solve this problem by applying a transfer-learning-based neural network approach to classify child and adult separately in camera images. The described work is comprised of image data collection, data annotation, transfer learning-based model development, and evaluation. For the work, Mask-RCNN (region-based convolutional neural network) with different backbone architectures and two different baselines are investigated and the perception precision of the architectures after transfer-learning is compared. The results reveal that the best performing trained model is able to detect and classify children and adults separately in different road scenarios with segmentation mask AP (average precision) of 85% and bounding box AP of 92%.
An AC-battery or multilevel inverter used to increase safety and flexibility is realizable by switching the cells and modules on and off in a defined way and thus can replace the bidirectional converter. Assessing possible additional aging due to switching, the results of a previous study for a high-power optimized cell showed no influence on the current rates or the switching. In this paper, a highly energy-optimized LG 18650-cell is investigated to discuss the influence of switching during the charge and discharge process, respectively, as well as combining both processes together with clear performance differences when applying higher charge and discharge currents. Moreover, the influence of switching is discussed for the two frequencies (50 Hz and 10 kHz) and different duty cycles. The aging is analyzed by capacity loss and resistance increase, by dV/dQ analysis, and by electrochemical impedance spectroscopy. We found no clear negative influence of switching but a positive effect if the cells are switched during charge. The best performance is found for switching during charge as well as during discharge. The cell aging during switching is clearly determined by the average and not the maximum current applied. This work shows no negative effects of multilevel inverter applications on the tested cells.