DeepGreen
Refine
Year of publication
- 2023 (17) (remove)
Document Type
- Article (17)
Language
- English (17)
Has Fulltext
- yes (17)
Keywords
- Neural Network (2)
- 3D Recovery (1)
- Albedo (1)
- All-Wheel Drive (1)
- Assistive Technologies (1)
- Automotive (1)
- Base-Catalyzed (1)
- Benutzererlebnis (1)
- Bestärkendes Lernen <Künstliche Intelligenz> (1)
- Bewältigung (1)
Faculty
- Angewandte Naturwissenschaften (F11) (6)
- Fakultät 12 / Institut für Technologie und Ressourcenmanagement in den Tropen und Subtropen (2)
- Fakultät 03 / Institut für Informationsmanagement (1)
- Fakultät 04 / Schmalenbach Institut für Wirtschaftswissenschaften (1)
- Fakultät 07 / Institut für Elektrische Energietechnik (1)
- Fakultät 07 / Institut für Medien- und Phototechnik (1)
- Fakultät 09 / Cologne Institute for Renewable Energy (1)
- Fakultät 09 / Institut für Produktentwicklung und Konstruktionstechnik (1)
- Fakultät 09 / Sonstige (1)
- Fakultät 10 / Institut Allgemeiner Maschinenbau (1)
Feasibility Study of Wheel Torque Prediction with a Recurrent Neural Network Using Vehicle Data
(2023)
In this paper, we present a feasibility study on predicting the torque signal of a passenger car with the help of a neural network. In addition, we analyze the possibility of using the proposed model structure for temperature prediction. This was carried out with a neural network, specifically a three-layer long short-term memory (LSTM) network. The data used were real road load data from a Jaguar Land Rover Evoque with a Twinster gearbox from GKN. The torque prediction generated good results with an accuracy of 55% and a root mean squared error (RMSE) of 49 Nm, considering that the data were not generated under laboratory conditions. However, the performance of predicting the temperature signal was not satisfying with a coefficient of determination (R2) score of −1.396 and an RMSE score of 69.4 °C. The prediction of the torque signal with the three-layer LSTM network was successful but the transferability of the network to another signal (temperature) was not proven. The knowledge gained from this investigation can be of importance for the development of virtual sensor technology.
The demand for explainable and transparent models increases with the continued success of reinforcement learning. In this article, we explore the potential of generating shallow decision trees (DTs) as simple and transparent surrogate models for opaque deep reinforcement learning (DRL) agents. We investigate three algorithms for generating training data for axis-parallel and oblique DTs with the help of DRL agents (“oracles”) and evaluate these methods on classic control problems from OpenAI Gym. The results show that one of our newly developed algorithms, the iterative training, outperforms traditional sampling algorithms, resulting in well-performing DTs that often even surpass the oracle from which they were trained. Even higher dimensional problems can be solved with surprisingly shallow DTs. We discuss the advantages and disadvantages of different sampling methods and insights into the decision-making process made possible by the transparent nature of DTs. Our work contributes to the development of not only powerful but also explainable RL agents and highlights the potential of DTs as a simple and effective alternative to complex DRL models.
Ten years after the journal’s first publication, we are taking a closer look at the knowledge flows of the output of the journal Publications. We analyzed the papers, topics, their authors and countries to assess the development of scholarly communication within Publications. Our bibliometric analyses show the research journal’s community, where the knowledge of this community is coming from, where it is going, and how diverse the community is based on its internationality and multidisciplinarity. We compare these findings with the scopes and topical goals the journal specifies. We aim at informing the editors and editorial board about the journal’s development to advance the journal’s role in scholarly communication. The results show that regarding topical diversity and internationality, the journal has remarkably developed. Moreover, the journal tends towards the field of library and information science, but strengthens its multidisciplinary status via its topics and author backgrounds.
Ground tire rubber (GTR) is a product obtained by grinding worn tire treads before retreading them or via the cryogenic or ambient temperature milling of end-of-life tires (ELTs). The aim of this study is to evaluate if calcium carbonate can be substituted by GTR and, if so, to what extent. Different types of ground tire rubber are incorporated in an EPDM (ethylene–propylene–diene–rubber) model compound as partial or complete substitutes of calcium carbonate. The raw compounds and the vulcanizates are characterized to identify the limits. In general, it is apparent that increasing amounts of GTR and larger particles degrade the mechanical properties. The GTR also influences the vulcanization kinetics by reducing the scorch time up to 50% and vulcanization time up to nearly 80%. This is significant for production processes. The compounds with one-third substitution with the smaller-particle-size GTR show mostly similar or even better properties than the reference.
Maximising Distribution Grid Utilisation by Optimising E-Car Charging Using Smart Meter Gateway Data
(2023)
The transition towards climate neutrality will result in an increase in electrical vehicles, as well as other electric loads, leading to higher loads on electrical distribution grids. This paper presents an optimisation algorithm that enables the integration of more loads into distribution grid infrastructure using information from smart meters and/or smart meter gateways. To achieve this, a mathematical programming formulation was developed and implemented. The algorithm determines the optimal charging schedule for all electric vehicles connected to the distribution grid, taking into account various criteria to avoid violating physical grid limitations and ensuring non-discriminatory charging of all electric vehicles on the grid while also optimising grid operation. Additionally, the expandability of the infrastructure and fail-safe operation are considered through the decentralisation of all components. Various scenarios are modelled and evaluated in a simulation environment. The results demonstrate that the developed optimisation algorithm allows for higher transformer loads compared to a P(U) control approach, without causing grid overload as observed in scenarios without optimisation or P(U) control.
Abstract
Even though researchers are increasingly acknowledging the dark side of customer participation (i.e., behavioral customer engagement), particularly in professional services with high cognitive demands that cause customer participation stress (i.e., negative psychological state resulting from the customer's overextension by required customer participation efforts), insights on how firms can effectively mitigate customer participation stress remains limited. Building on transactional stress theory, we investigate whether customers can effectively cope with the expected cognitive demands of professional services. Moreover, by introducing an adapted coping construct (i.e., coping support), we examine whether employees can provide coping support to help decrease customer participation stress. The findings of a time‐lagged field study with customers of a large German bank (N = 117) suggest that customer coping before the encounter cannot mitigate the effect of anticipated cognitive demands on customer participation stress. Instead, the results of both the field study and a follow‐up experimental study (N = 218) show that a certain set of employee coping support during service encounters is crucial. While focusing on action coping support is not ideal in situations with high cognitive demands, firms should advise their professional service employees to offer emotional coping support to attenuate the unfavorable effect of cognitive demands on customer participation stress.
To date, the establishment of high-titer stable viral packaging cells (VPCs) at large scale for gene therapeutic applications is very time- and cost-intensive. Here we report the establishment of three human suspension 293-F-derived ecotropic MLV-based VPCs. The classic stable transfection of an EGFP-expressing transfer vector resulted in a polyclonal VPC pool that facilitated cultivation in shake flasks of 100 mL volumes and yielded high functional titers of more than 1 × 106 transducing units/mL (TU/mL). When the transfer vector was flanked by transposon terminal inverted repeats (TIRs) and upon co-transfection of a plasmid encoding for the transposase, productivities could be slightly elevated to more than 3 × 106 TU/mL. In contrast and using mRNA encoding for the transposase, as a proof of concept, productivities were drastically improved by more than ten-fold exceeding 5 × 107 TU/mL. In addition, these VPC pools were generated within only 3 weeks. The production volume was successfully scaled up to 500 mL employing a stirred-tank bioreactor (STR). We anticipate that the stable transposition of transfer vectors employing transposase transcripts will be of utility for the future establishment of high-yield VPCs producing pseudotype vector particles with a broader host tropism on a large scale.
Austria is committed to the net-zero climate goal along with the European Union. This requires all sectors to be decarbonized. Hereby, hydrogen plays a vital role as stated in the national hydrogen strategy. A report commissioned by the Austrian government predicts a minimum hydrogen demand of 16 TWh per year in Austria in 2040. Besides hydrogen imports, domestic production can ensure supply. Hence, this study analyses the levelized cost of hydrogen for an off-grid production plant including a proton exchange membrane electrolyzer, wind power and solar photovoltaics in Austria. In the first step, the capacity factors of the renewable electricity sources are determined by conducting a geographic information system analysis. Secondly, the levelized cost of electricity for wind power and solarphotovoltaics plants in Austria is calculated. Thirdly, the most cost-efficient portfolio of wind power and solar photovoltaics plants is determined using electricity generation profiles with a 10-min granularity. The modelled system variants differ among location, capacity factors of the renewable electricity sources and the full load hours of the electrolyzer. Finally, selected variables are tested for their sensitivities. With the applied model, the hydrogen production cost for decentralized production plants can be calculated for any specific location. The levelized cost of hydrogen estimates range from 3.08 EUR/kg to 13.12 EUR/kg of hydrogen, whereas it was found that the costs are most sensitive to the capacity factors of the renewable electricity sources and the full load hours of the electrolyzer. The novelty of the paper stems from the model applied that calculates the levelized cost of renewable hydrogen in an off-grid hydrogen production system. The model finds a cost-efficient portfolio of directly coupled wind power and solar photovoltaics systems for 80 different variants in an Austria-specific context.
In this work, we propose a novel data-driven approach to recover missing or corrupted motion capture data, either in the form of 3D skeleton joints or 3D marker trajectories. We construct a knowledge-base that contains prior existing knowledge, which helps us to make it possible to infer missing or corrupted information of the motion capture data. We then build a kd-tree in parallel fashion on the GPU for fast search and retrieval of this already available knowledge in the form of nearest neighbors from the knowledge-base efficiently. We exploit the concept of histograms to organize the data and use an off-the-shelf radix sort algorithm to sort the keys within a single processor of GPU. We query the motion missing joints or markers, and as a result, we fetch a fixed number of nearest neighbors for the given input query motion. We employ an objective function with multiple error terms that substantially recover 3D joints or marker trajectories in parallel on the GPU. We perform comprehensive experiments to evaluate our approach quantitatively and qualitatively on publicly available motion capture datasets, namely CMU and HDM05. From the results, it is observed that the recovery of boxing, jumptwist, run, martial arts, salsa, and acrobatic motion sequences works best, while the recovery of motion sequences of kicking and jumping results in slightly larger errors. However, on average, our approach executes outstanding results. Generally, our approach outperforms all the competing state-of-the-art methods in the most test cases with different action sequences and executes reliable results with minimal errors and without any user interaction.
As the population ages, the demand for care for older adults is increasing. To maintain their independence and autonomy, even with declining health, assistive technologies such as connected medical devices or social robots can be useful. In previous work, we introduced a novel health monitoring system that combines commercially available products with apps designed specifically for older adults. The system is intended for the long-term collection of subjective and objective health data. In this work, we present an exploratory user experience (UX) and usability study we conducted with older adults as the target group of the system and with younger expert users who tested our msystem. All participants interacted with a social robot conducting a health assessment and tested sensing devices and an app for data visualization. The UX and usability of the individual components of the system were rated highly in questionnaires in all sessions. All participants also said they would use such a system in their everyday lives, demonstrating the potential of these systems for self-managing users’ health. Finally, we found factors such as previous experience with social robots and technological expertise to have an influence on the reported UX of the users.
During spaceflight, humans experience a variety of physiological changes due to deviations from familiar earth conditions. Specifically, the lack of gravity is responsible for many effects observed in returning astronauts. These impairments can include structural as well as functional changes of the brain and a decline in cognitive performance. However, the underlying physiological mechanisms remain elusive. Alterations in neuronal activity play a central role in mental disorders and altered neuronal transmission may also lead to diminished human performance in space. Thus, understanding the influence of altered gravity at the cellular and network level is of high importance. Previous electrophysiological experiments using patch clamp techniques and calcium indicators have shown that neuronal activity is influenced by altered gravity. By using multi-electrode array (MEA) technology, we advanced the electrophysiological investigation covering single-cell to network level responses during exposure to decreased (micro-) or increased (hyper-) gravity conditions. We continuously recorded in real-time the spontaneous activity of human induced pluripotent stem cell (hiPSC)-derived neural networks in vitro. The MEA device was integrated into a custom-built environmental chamber to expose the system with neuronal cultures to up to 6 g of hypergravity on the Short-Arm Human Centrifuge at the DLR Cologne, Germany. The flexibility of the experimental hardware set-up facilitated additional MEA electrophysiology experiments under 4.7 s of high-quality microgravity (10–6 to 10–5 g) in the Bremen drop tower, Germany. Hypergravity led to significant changes in activity. During the microgravity phase, the mean action potential frequency across the neural networks was significantly enhanced, whereas different subgroups of neurons showed distinct behaviors, such as increased or decreased firing activity. Our data clearly demonstrate that gravity as an environmental stimulus triggers changes in neuronal activity. Neuronal networks especially reacted to acute changes in mechanical loading (hypergravity) or de-loading (microgravity). The current study clearly shows the gravity-dependent response of neuronal networks endorsing the importance of further investigations of neuronal activity and its adaptive responses to micro- and hypergravity. Our approach provided the basis for the identification of responsible mechanisms and the development of countermeasures with potential implications on manned space missions.
The European heating sector is currently heavily dominated by fossil fuels. Composting is a naturally occurring process in which heat is liberated from the composting substrate at a higher rate than the process needs to support itself. This difference could be harnessed for low-heat applications such as residential consumption, alleviating some of the impacts fossil fuel emissions represent. In this study, the composting heat recovery reported in the literature was compared to the energy demand for space and water heating in four European countries. A review of potential heat production from the waste representative of the residential sector was performed. We found that the theoretically recoverable composting heat does not significantly reduce the need for district heating. However, it can significantly reduce the energy demand for water heating, being able to supply countries such as Greece with between 36% and 100% of the yearly hot water demand, or 12% to 53% of the yearly hot water of countries such as Switzerland, depending on the efficiency of heat recovery.
The Production of Isophorone
(2023)
Isophorone is a technically important compound used as a high-boiling-point solvent for coatings, adhesives, etc., and it is used as a starting material for various valuable compounds, including isophorone diisocyanate, a precursor for polyurethanes. For over 80 years, isophorone has been synthesized via base-catalyzed self-condensation of acetone. This reaction has a complex reaction mechanism with numerous possible reaction steps including the formation of isophorone, triacetone dialcohol, and ketonic resins. This review provides an overview of the different production processes of isophorone in liquid- and vapor-phase and reviews the literature-reported selectivity toward isophorone achieved using different reaction parameters and catalysts.
The annual yield of bifacial photovoltaic systems is highly dependent on the albedo of the underlying soil. There are currently no published data about the albedo of red soil in western Africa. In this study, the impact of the albedo of red soil in Ghana on the energy yield of bifacial photovoltaic systems is analysed. A bifacial photovoltaic simulation model is created by combining the optical view factor matrix with an electrical output simulation. For an exact simulation, the albedo of red soil at three different locations in Ghana is measured for the first time. The average albedo of every red soil is clearly determined, as well as the measurement span including instrumentation uncertainty; values between 0.175 and 0.335 were measured. Considering these data, a state-of-the-art bifacial photovoltaic system with an average of 19.8% efficient modules in northern Ghana can achieve an annual energy yield of 508.8 kWh/m2 and a bifacial gain of up to 18.3% in comparison with monofacial photovoltaic panels. To summarise, red soil in two out of three locations in Ghana shows higher albedo values than most natural ground surfaces and therefore positively impacts the annual yield of bifacial photovoltaic systems.
Porous polymer membranes substantially contribute to an acceleration of sustainability transformation based on the energy efficient separation of liquid and gaseous mixtures. This rapid shift toward sustainable industrial processes leads to an increased demand for specifically tailored membranes. In order to predict membrane performance factors like permeability, selectivity and durability, the membrane formation process by film casting and phase inversion needs to be understood further. In recent years, computational models of the membrane formation process have been studied intensely. Their high spatial and temporal resolution allows a detailed quantitative description of phase inversion phenomena. New experimental techniques complement this development, as they provide quantitative data, e.g., on compositional changes of the polymer solution during membrane formation as well as the kinetic progression of the phase separation process. This state-of-the-art review compiles computational and experimental approaches that characterize the phase inversion process. We discuss how this methodological pluralism is necessary for improving the tailoring of membrane parameters, but that it is unlikely to be the way to the ultimate goal of a complete description of the evolution of the membrane structure from the initial demixing to the final solidification. Alternatively, we formulate an approach that includes a database of standardized and harmonized membrane performance data based on previously publicized data, as well as the application of artificial neural networks as a new powerful tool to link membrane production parameters to membrane performance.
Agents with antifungal activity play a vital role as therapeutics in health care, as do fungicides in agriculture. Effectiveness, toxicological profile, and eco-friendliness are among the properties used to select suitable substances. Furthermore, a steady supply of new agents with different modes of action is required to counter the well-known potential of human and phyto-pathogenic fungi to develop resistance against established antifungals. Here, we use an in vitro growth assay to investigate the activity of the calcineurin inhibitor tacrolimus in combination with the commercial fungicides cyproconazole and hymexazol, as well as with two earlier reported novel {2-(3-R-1H-1,2,4-triazol-5-yl)phenyl}amines, against the fungi Aspergillus niger, Colletotrichum higginsianum, Fusarium oxysporum and the oomycete Phytophthora infestans, which are notoriously harmful in agriculture. When tacrolimus was added in a concentration range from 0.25 to 25 mg/L to the tested antifungals (at a fixed concentration of 25 or 50 mg/L), the inhibitory activities were distinctly enhanced. Molecular docking calculations revealed triazole derivative 5, (2-(3-adamantan-1-yl)-1H-1,2,4-triazol-5-yl)-4-chloroaniline), as a potent inhibitor of chitin deacetylases (CDA) of Aspergillus nidulans and A. niger (AnCDA and AngCDA, respectively), which was stronger than the previously reported polyoxorin D, J075-4187, and chitotriose. The results are discussed in the context of potential synergism and molecular mode of action.
Hydroxybenzene, commonly known as phenol, is one of the most important organic commodity chemicals. To produce phenol, the cumene process is the most used process worldwide. A crucial step in this process is the Hock rearrangement, which has a major impact on the overall cumene consumption rate and determines the safety level of the process. The most used catalyst for the cleavage of cumene hydroperoxide (CHP) is sulfuric acid. Besides its strong corrosive property, which increases plant investment costs, it also requires neutralization after the decomposition step to prevent side reactions. In this study, we show that high-temperature-treated Linde Type X (LTX) zeolites exhibit a high activity for the peroxide cleavage step. In addition, the structure–activity relationship responsible for this good performance in the reaction system of the HOCK rearrangement was investigated. XRPD analyses revealed the formation of a new phase after temperature treatment above 900 °C. The Si/Al ratio determined by EDX suggested the formation of extra-framework aluminum, which was confirmed by solid-state NMR analysis. The newly formed extra-framework aluminum was found to be responsible for the high catalytic activity. BET analyses showed that the surface area drops at higher calcination temperatures. This leads to a lower catalytic activity for most known reactions. However, for this study, no decrease in activity has been observed. The newfound material shows extraordinarily high activity as a catalyst in the HOCK cleavage and has the potential to be a heterogeneous alternative to sulfuric acid for this reaction.