Search

Wind turbines (WTs) are loaded by the wind and by the electrical grid. A low voltage ride through (LVRT) is a special load case for WTs, which can be critical for their drivetrain. They are usually less severe for WTs with a full-scale converter (FSC); however, in some cases there are still loads observable in the mechanical drivetrains of WTs with FSCs. Different aspects like the brake chopper design, the control and specifications of the grid side converter (GSC) as well as the grid and fault parameters affect these loads. In this paper, three different LVRT behaviors of a WT with a squirrel-cage induction generator (SCIG) and a FSC were investigated via simulations. In the worst case, a sharp peak in the electromagnetic torque of the generator after fault clearance is possible. The cause lies in the GSC control, especially the phase-locked loop (PLL), and can also be influenced by adjusting the control parameters, brake chopper and filter design. The resulting mechanical drivetrain torque, on the other hand, is increased by only 2.9% for a very short time. Thus, the loading of the mechanical drivetrain components is only slightly increased. Therefore, the risk of damage to the mechanical components of the FSC WT due to grid faults is relatively low.

Smart City applications such as intelligent traffic routing or accident prevention rely on computer vision methods for exact vehicle localization and tracking. Due to the scarcity of accurately labeled data, detecting and tracking vehicles in 3D from multiple cameras proves challenging to explore. We present a massive synthetic dataset for multiple vehicle tracking and segmentation in multiple overlapping and non-overlapping camera views. Unlike existing datasets, which only provide tracking ground truth for 2D bounding boxes, our dataset additionally contains perfect labels for 3D bounding boxes in camera- and world coordinates, depth estimation, and instance, semantic and panoptic segmentation. The dataset consists of 17 hours of labeled video material, recorded from 340 cameras in 64 diverse day, rain, dawn, and night scenes, making it the most extensive dataset for multi-target multi-camera tracking so far. We provide baselines for detection, vehicle re-identification, and single- and multi-camera tracking. Code and data are publicly available.

Person Re-Identification aims to retrieve person identities from images captured by multiple cameras or the same cameras in different time instances and locations. Because of its importance in many vision applications from surveillance to human-machine interaction, person re-identification methods need to be reliable and fast. While more and more deep architectures are proposed for increasing performance, those methods also increase overall model complexity. This paper proposes a lightweight network that combines global, part-based, and channel features in a unified multi-branch architecture that builds on the resource-efficient OSNet backbone. Using a well-founded combination of training techniques and design choices, our final model achieves state-of-the-art results on CUHK03 labeled, CUHK03 detected, and Market-1501 with 85.1% mAP / 87.2% rank1, 82.4% mAP / 84.9% rank1, and 91.5% mAP / 96.3% rank1, respectively., Comment: 5 pages, 1 figure

Gait recognition is a promising video-based biometric for identifying individual walking patterns from a long distance. At present, most gait recognition methods use silhouette images to represent a person in each frame. However, silhouette images can lose fine-grained spatial information, and most papers do not regard how to obtain these silhouettes in complex scenes. Furthermore, silhouette images contain not only gait features but also other visual clues that can be recognized. Hence these approaches can not be considered as strict gait recognition. We leverage recent advances in human pose estimation to estimate robust skeleton poses directly from RGB images to bring back model-based gait recognition with a cleaner representation of gait. Thus, we propose GaitGraph that combines skeleton poses with Graph Convolutional Network (GCN) to obtain a modern model-based approach for gait recognition. The main advantages are a cleaner, more elegant extraction of the gait features and the ability to incorporate powerful spatio-temporal modeling using GCN. Experiments on the popular CASIA-B gait dataset show that our method archives state-of-the-art performance in model-based gait recognition. The code and models are publicly available., 5 pages, 2 figures

Understanding actions and gestures in video streams requires temporal reasoning of the spatial content from different time instants, i.e., spatiotemporal (ST) modeling. In this survey paper, we have made a comparative analysis of different ST modeling techniques for action and gesture recognition tasks. Since Convolutional Neural Networks (CNNs) are proved to be an effective tool as a feature extractor for static images, we apply ST modeling techniques on the features of static images from different time instants extracted by CNNs. All techniques are trained end-to-end together with a CNN feature extraction part and evaluated on two publicly available benchmarks: The Jester and the Something-Something datasets. The Jester dataset contains various dynamic and static hand gestures, whereas the Something-Something dataset contains actions of human-object interactions. The common characteristic of these two benchmarks is that the designed architectures need to capture the full temporal content of videos in order to correctly classify actions/gestures. Contrary to expectations, experimental results show that Recurrent Neural Network (RNN) based ST modeling techniques yield inferior results compared to other techniques such as fully convolutional architectures. Codes and pretrained models of this work are publicly available (https://github.com/fubel/stmodeling).

Forschung im Ingenieurwesen : research journal = Engineering research (2021). doi:10.1007/s10010-021-00461-2, Published by Springer, Berlin ; Heidelberg

Doubly fed induction generators are widely used in wind turbines in the field. The main advantage is the significantly lower cost of the partial scale power converter compared to the full scale power converter. Converter faults in wind turbines lead to significant generator torque excitations that lead to dynamic loads in the drivetrain. Dynamic loads and changing rotational speeds can lead to gearbox damages. The generator torque excitations have the highest influence on the high speed shaft torque due to the coupling to the generator. Gearbox damages occur mainly on the components of the high speed shaft. Thus, in this paper the influence of a converter fault in a wind turbine with doubly fed induction generator on the high speed shaft component damages is investigated. It is shown that the converter fault can induce dynamic torque excitations with a maximum increase to around 2.5 times rated torque. Due to the resulting dynamic gearbox loading the safety against scuffing in the high speed gear stage decreases by maximum 19 percent. The smearing risk in the high speed shaft bearings increases to around 2 times the value during rated power production.

Convolutional Neural Networks with 3D kernels (3D CNNs) currently achieve state-of-the-art results in video recognition tasks due to their supremacy in extracting spatiotemporal features within video frames. There have been many successful 3D CNN architectures surpassing the state-of-the-art results successively. However, nearly all of them are designed to operate offline creating several serious handicaps during online operation. Firstly, conventional 3D CNNs are not dynamic since their output features represent the complete input clip instead of the most recent frame in the clip. Secondly, they are not temporal resolution-preserving due to their inherent temporal downsampling. Lastly, 3D CNNs are constrained to be used with fixed temporal input size limiting their flexibility. In order to address these drawbacks, we propose dissected 3D CNNs, where the intermediate volumes of the network are dissected and propagated over depth (time) dimension for future calculations, substantially reducing the number of computations at online operation. For action classification, the dissected version of ResNet models performs 74-90% fewer computations at online operation while achieving $\sim$5% better classification accuracy on the Kinetics-600 dataset than conventional 3D ResNet models. Moreover, the advantages of dissected 3D CNNs are demonstrated by deploying our approach onto several vision tasks, which consistently improved the performance.

After binding to its cell surface receptor ganglioside GM1, simian virus 40 (SV40) is endocytosed by lipid raft-mediated endocytosis and slowly transported to the endoplasmic reticulum, where partial uncoating occurs. We analyzed the intracellular pathway taken by the virus in HeLa and CV-1 cells by using a targeted small interfering RNA (siRNA) silencing screen, electron microscopy, and live-cell imaging as well as by testing a variety of cellular inhibitors and other perturbants. We found that the virus entered early endosomes, late endosomes, and probably endolysosomes before reaching the endoplasmic reticulum and that this pathway was part of the infectious route. The virus was especially sensitive to a variety of perturbations that inhibited endosome acidification and maturation. Contrary to our previous models, which postulated the passage of the virus through caveolin-rich organelles that we called caveosomes, we conclude that SV40 depends on the classical endocytic pathway for infectious entry.

In the emerging market of nano-sized products, silver nanoparticles (Ag NPs) are widely used due to their antimicrobial properties. Human interaction with Ag NPs can occur through the lung, skin, gastrointestinal tract, and bloodstream. However, the inhalation of Ag NP aerosols is a primary concern. To study the possible effects of inhaled Ag NPs, an in vitro triple cell co-culture model of the human alveolar/airway barrier (A549 epithelial cells, human peripheral blood monocyte derived dendritic and macrophage cells) together with an air–liquid interface cell exposure (ALICE) system was used in order to reflect a real-life exposure scenario. Cells were exposed at the air–liquid interface (ALI) to 0.03, 0.3, and 3 µg Ag/cm2 of Ag NPs (diameter 100 nm; coated with polyvinylpyrrolidone: PVP). Ag NPs were found to be highly aggregated within ALI exposed cells with no impairment of cell morphology. Furthermore, a significant increase in release of cytotoxic (LDH), oxidative stress (SOD-1, HMOX-1) or pro-inflammatory markers (TNF-α, IL-8) was absent. As a comparison, cells were exposed to Ag NPs in submerged conditions to 10, 20, and 30 µg Ag/mL. The deposited dose per surface area was estimated by using a dosimetry model (ISDD) to directly compare submerged vs ALI exposure concentrations after 4 and 24 h. Unlike ALI exposures, the two highest concentrations under submerged conditions promoted a cytotoxic and pro-inflammatory response after 24 h. Interestingly, when cell cultures were co-incubated with lipopolysaccharide (LPS), no synergistic inflammatory effects were observed. By using two different exposure scenarios it has been shown that the ALI as well as the suspension conditions for the lower concentrations after 4 h, reflecting real-life concentrations of an acute 24 h exposure, did not induce any adverse effects in a complex 3D model mimicking the human alveolar/airway barrier. However, the highest concentrations used in the ALI setup, as well as all concentrations under submerged conditions after 24 h, reflecting more of a chronic lifetime exposure concentration, showed cytotoxic as well as pro-inflammatory effects. In conclusion, more studies need to address long-term and chronic Ag NP exposure effects.

Owing to their antimicrobial properties, silver nanoparticles (NPs) are the most commonly used engineered nanomaterial for use in a wide array of consumer and medical applications. Many discussions are currently ongoing as to whether or not exposure of silver NPs to the ecosystem (i.e. plants and animals) may be conceived as harmful or not. Metallic silver, if released into the environment, can undergo chemical and biochemical conversion which strongly influence its availability towards any biological system. During this process, in the presence of moisture, silver can be oxidized resulting in the release of silver ions. To date, it is still debatable as to whether any biological impact of nanosized silver is relative to either its size, or to its ionic constitution. The aim of this review therefore is to provide a comprehensive, interdisciplinary overview—for biologists, chemists, toxicologists as well as physicists—regarding the production of silver NPs, its (as well as in their ionic form) chemical and biochemical behaviours towards/within a multitude of relative and realistic biological environments and also how such interactions may be correlated across a plethora of different biological organisms.

How non-enveloped viruses overcome host cell membranes is poorly understood. Here, we show that after endocytosis and transport to the endoplasmic reticulum (ER), but before crossing the ER membrane to the cytosol, incoming simian virus 40 particles are structurally remodelled leading to exposure of the amino-terminal sequence of the minor viral protein VP2. These hydrophobic sequences anchor the virus to membranes. A negatively charged residue, Glu 17, in the α-helical, membrane-embedded peptide is essential for infection, most likely by introducing an 'irregularity' recognized by the ER-associated degradation (ERAD) system for membrane proteins. Using a siRNA-mediated screen, the lumenal chaperone BiP and the ER-membrane protein BAP31 (both involved in ERAD) were identified as being essential for infection. They co-localized with the virus in discrete foci and promoted its ER-to-cytosol dislocation. Virus-like particles devoid of VP2 failed to cross the membrane. The results demonstrated that ERAD-factors assist virus transport across the ER membrane.

There are many studies with cells to find out how particles interact with them. In contrast to micronsized particles, which are actively taken up by phagocytosis or macropinocytosis, nanosized particles may be taken up by cells through different endocytic pathways or by another, yet to be defined mechanism. There is increasing evidence that it is the nanosized particles, which are a particular risk because of their high content of organic chemicals and their pro-oxidative potential due to the high surface-to-volume ratio of the particles as compared to the bulk material. It is the goal of this article to create an understanding for the interaction of particles with biological systems, with particular consideration of the interaction of nanoparticles (NPs) with lung cells. One is attempting to understand, how NPs interact with cellular membranes, as it is hardly known, how they are taken up by cells, how they are trafficking in cells, and how they interact with subcellular compartments, such as with mitochondria or with the nucleus. Cells tend to defend themselves against any foreign material, which is taken up. In general, they try to eliminate particulate intruders and this is what they usually manage with micronsized particles. However, with NPs it is different. NPs may not be eliminated easily, and, hence may stimulate the cells to react in an unfavorable way. What we can learn is that NPs behave differently than microparticles.

After binding to its cell surface receptor ganglioside GM1, simian virus 40 (SV40) is endocytosed by lipid raft-mediated endocytosis and slowly transported to the endoplasmic reticulum, where partial uncoating occurs. We analyzed the intracellular pathway taken by the virus in HeLa and CV-1 cells by using a targeted small interfering RNA (siRNA) silencing screen, electron microscopy, and live-cell imaging as well as by testing a variety of cellular inhibitors and other perturbants. We found that the virus entered early endosomes, late endosomes, and probably endolysosomes before reaching the endoplasmic reticulum and that this pathway was part of the infectious route. The virus was especially sensitive to a variety of perturbations that inhibited endosome acidification and maturation. Contrary to our previous models, which postulated the passage of the virus through caveolin-rich organelles that we called caveosomes, we conclude that SV40 depends on the classical endocytic pathway for infectious entry.

PVP-capped silver nanoparticles with a diameter of the metallic core of 70 nm, a hydrodynamic diameter of 120 nm and a zeta potential of −20 mV were prepared and investigated with regard to their biological activity. This review summarizes the physicochemical properties (dissolution, protein adsorption, dispersability) of these nanoparticles and the cellular consequences of the exposure of a broad range of biological test systems to this defined type of silver nanoparticles. Silver nanoparticles dissolve in water in the presence of oxygen. In addition, in biological media (i.e., in the presence of proteins) the surface of silver nanoparticles is rapidly coated by a protein corona that influences their physicochemical and biological properties including cellular uptake. Silver nanoparticles are taken up by cell-type specific endocytosis pathways as demonstrated for hMSC, primary T-cells, primary monocytes, and astrocytes. A visualization of particles inside cells is possible by X-ray microscopy, fluorescence microscopy, and combined FIB/SEM analysis. By staining organelles, their localization inside the cell can be additionally determined. While primary brain astrocytes are shown to be fairly tolerant toward silver nanoparticles, silver nanoparticles induce the formation of DNA double-strand-breaks (DSB) and lead to chromosomal aberrations and sister-chromatid exchanges in Chinese hamster fibroblast cell lines (CHO9, K1, V79B). An exposure of rats to silver nanoparticles in vivo induced a moderate pulmonary toxicity, however, only at rather high concentrations. The same was found in precision-cut lung slices of rats in which silver nanoparticles remained mainly at the tissue surface. In a human 3D triple-cell culture model consisting of three cell types (alveolar epithelial cells, macrophages, and dendritic cells), adverse effects were also only found at high silver concentrations. The silver ions that are released from silver nanoparticles may be harmful to skin with disrupted barrier (e.g., wounds) and induce oxidative stress in skin cells (HaCaT). In conclusion, the data obtained on the effects of this well-defined type of silver nanoparticles on various biological systems clearly demonstrate that cell-type specific properties as well as experimental conditions determine the biocompatibility of and the cellular responses to an exposure with silver nanoparticles.

Background Due to its antibacterial properties, silver (Ag) has been used in more consumer products than any other nanomaterial so far. Despite the promising advantages posed by using Ag-nanoparticles (NPs), their interaction with mammalian systems is currently not fully understood. An exposure route via inhalation is of primary concern for humans in an occupational setting. Aim of this study was therefore to investigate the potential adverse effects of aerosolised Ag-NPs using a human epithelial airway barrier model composed of A549, monocyte derived macrophage and dendritic cells cultured in vitro at the air-liquid interface. Cell cultures were exposed to 20 nm citrate-coated Ag-NPs with a deposition of 30 and 278 ng/cm2 respectively and incubated for 4 h and 24 h. To elucidate whether any effects of Ag-NPs are due to ionic effects, Ag-Nitrate (AgNO3) solutions were aerosolised at the same molecular mass concentrations. Results Agglomerates of Ag-NPs were detected at 24 h post exposure in vesicular structures inside cells but the cellular integrity was not impaired upon Ag-NP exposures. Minimal cytotoxicity, by measuring the release of lactate dehydrogenase, could only be detected following a higher concentrated AgNO3-solution. A release of pro-inflammatory markers TNF-α and IL-8 was neither observed upon Ag-NP and AgNO3 exposures as well as was not affected when cells were pre-stimulated with lipopolysaccharide (LPS). Also, an induction of mRNA expression of TNF-α and IL-8, could only be observed for the highest AgNO3 concentration alone or even significantly increased when pre-stimulated with LPS after 4 h. However, this effect disappeared after 24 h. Furthermore, oxidative stress markers (HMOX-1, SOD-1) were expressed after 4 h in a concentration dependent manner following AgNO3 exposures only. Conclusions With an experimental setup reflecting physiological exposure conditions in the human lung more realistic, the present study indicates that Ag-NPs do not cause adverse effects and cells were only sensitive to high Ag-ion concentrations. Chronic exposure scenarios however, are needed to reveal further insight into the fate of Ag-NPs after deposition and cell interactions.

In light of the increasingly abundant use of engineered nanoparticles(NPs) and the ongoing exposure to ambient ultrafine particles it is imperative that the potential for NPs to elicit adverse effects on human health is understood. In order to determine the potential harm that NPs may exert, many different in vitro systems have been used. Commonly in vitro nanotoxicology studies use NP suspensions applied directly to cell cultures. Although the use of in vitro monoculture systems to assess the effects of NPs on, for example, the lung is frequently debated, the use of suspension exposures is not realistic in relation to the exposure of NPs to humans via inhalation; the primary exposure route to the human body for NPs. As an alternative to the suspension (or submerged) exposure method, numerous different exposure systems at theair-liquid interface have been developed and used in nanotoxicology research, which mimic the realistic conditions of NP inhalation exposure. In addition, such air-liquid exposure systems also offer the advantage to determine the exact dose (or concentration) which is deposited on the cell surface. The aimof this review is to provide a description of these different exposure systems, to explain how they recreate realistic inhalation conditions for occupational and environmental exposure, as well to describe how they may be used to gain an insight into how NPs may interact with the epithelial airway barrier following inhalation.