@inproceedings{FreyerThewesMeinecke2023, author = {Freyer, Nils and Thewes, Dustin and Meinecke, Matthias}, title = {GUIDO: a hybrid approach to guideline discovery \& ordering from natural language texts}, series = {Proceedings of the 12th International Conference on Data Science, Technology and Applications DATA - Volume 1}, booktitle = {Proceedings of the 12th International Conference on Data Science, Technology and Applications DATA - Volume 1}, editor = {Gusikhin, Oleg and Hammoudi, Slimane and Cuzzocrea, Alfredo}, isbn = {978-989-758-664-4}, issn = {2184-285X}, doi = {10.5220/0012084400003541}, pages = {335 -- 342}, year = {2023}, abstract = {Extracting workflow nets from textual descriptions can be used to simplify guidelines or formalize textual descriptions of formal processes like business processes and algorithms. The task of manually extracting processes, however, requires domain expertise and effort. While automatic process model extraction is desirable, annotating texts with formalized process models is expensive. Therefore, there are only a few machine-learning-based extraction approaches. Rule-based approaches, in turn, require domain specificity to work well and can rarely distinguish relevant and irrelevant information in textual descriptions. In this paper, we present GUIDO, a hybrid approach to the process model extraction task that first, classifies sentences regarding their relevance to the process model, using a BERT-based sentence classifier, and second, extracts a process model from the sentences classified as relevant, using dependency parsing. The presented approach achieves significantly better resul ts than a pure rule-based approach. GUIDO achieves an average behavioral similarity score of 0.93. Still, in comparison to purely machine-learning-based approaches, the annotation costs stay low.}, language = {en} } @inproceedings{KloeserBuesgenKohletal.2023, author = {Kl{\"o}ser, Lars and B{\"u}sgen, Andr{\´e} and Kohl, Philipp and Kraft, Bodo and Z{\"u}ndorf, Albert}, title = {Explaining relation classification models with semantic extents}, series = {Deep Learning Theory and Applications}, booktitle = {Deep Learning Theory and Applications}, editor = {Conte, Donatello and Fred, Ana and Gusikhin, Oleg and Sansone, Carlo}, publisher = {Springer}, address = {Cham}, isbn = {978-3-031-39058-6 (Print)}, doi = {10.1007/978-3-031-39059-3_13}, pages = {189 -- 208}, year = {2023}, abstract = {In recent years, the development of large pretrained language models, such as BERT and GPT, significantly improved information extraction systems on various tasks, including relation classification. State-of-the-art systems are highly accurate on scientific benchmarks. A lack of explainability is currently a complicating factor in many real-world applications. Comprehensible systems are necessary to prevent biased, counterintuitive, or harmful decisions. We introduce semantic extents, a concept to analyze decision patterns for the relation classification task. Semantic extents are the most influential parts of texts concerning classification decisions. Our definition allows similar procedures to determine semantic extents for humans and models. We provide an annotation tool and a software framework to determine semantic extents for humans and models conveniently and reproducibly. Comparing both reveals that models tend to learn shortcut patterns from data. These patterns are hard to detect with current interpretability methods, such as input reductions. Our approach can help detect and eliminate spurious decision patterns during model development. Semantic extents can increase the reliability and security of natural language processing systems. Semantic extents are an essential step in enabling applications in critical areas like healthcare or finance. Moreover, our work opens new research directions for developing methods to explain deep learning models.}, language = {en} } @inproceedings{KetelhutGoellBraunsteinetal.2019, author = {Ketelhut, Maike and G{\"o}ll, Fabian and Braunstein, Bjoern and Albracht, Kirsten and Abel, Dirk}, title = {Iterative learning control of an industrial robot for neuromuscular training}, series = {2019 IEEE Conference on Control Technology and Applications}, booktitle = {2019 IEEE Conference on Control Technology and Applications}, publisher = {IEEE}, address = {New York}, isbn = {978-1-7281-2767-5 (ePub)}, doi = {10.1109/CCTA.2019.8920659}, pages = {7 Seiten}, year = {2019}, abstract = {Effective training requires high muscle forces potentially leading to training-induced injuries. Thus, continuous monitoring and controlling of the loadings applied to the musculoskeletal system along the motion trajectory is required. In this paper, a norm-optimal iterative learning control algorithm for the robot-assisted training is developed. The algorithm aims at minimizing the external knee joint moment, which is commonly used to quantify the loading of the medial compartment. To estimate the external knee joint moment, a musculoskeletal lower extremity model is implemented in OpenSim and coupled with a model of an industrial robot and a force plate mounted at its end-effector. The algorithm is tested in simulation for patients with varus, normal and valgus alignment of the knee. The results show that the algorithm is able to minimize the external knee joint moment in all three cases and converges after less than seven iterations.}, language = {en} } @inproceedings{MaurerMiskiwAcostaetal.2023, author = {Maurer, Florian and Miskiw, Kim K. and Acosta, Rebeca Ramirez and Harder, Nick and Sander, Volker and Lehnhoff, Sebastian}, title = {Market abstraction of energy markets and policies - application in an agent-based modeling toolbox}, series = {EI.A 2023: Energy Informatics}, booktitle = {EI.A 2023: Energy Informatics}, editor = {Jorgensen, Bo Norregaard and Pereira da Silva, Luiz Carlos and Ma, Zheng}, publisher = {Springer}, address = {Cham}, isbn = {978-3-031-48651-7 (Print)}, doi = {10.1007/978-3-031-48652-4_10}, pages = {139 -- 157}, year = {2023}, abstract = {In light of emerging challenges in energy systems, markets are prone to changing dynamics and market design. Simulation models are commonly used to understand the changing dynamics of future electricity markets. However, existing market models were often created with specific use cases in mind, which limits their flexibility and usability. This can impose challenges for using a single model to compare different market designs. This paper introduces a new method of defining market designs for energy market simulations. The proposed concept makes it easy to incorporate different market designs into electricity market models by using relevant parameters derived from analyzing existing simulation tools, morphological categorization and ontologies. These parameters are then used to derive a market abstraction and integrate it into an agent-based simulation framework, allowing for a unified analysis of diverse market designs. Furthermore, we showcase the usability of integrating new types of long-term contracts and over-the-counter trading. To validate this approach, two case studies are demonstrated: a pay-as-clear market and a pay-as-bid long-term market. These examples demonstrate the capabilities of the proposed framework.}, language = {en} } @inproceedings{SchulzeBuxlohGross2021, author = {Schulze-Buxloh, Lina and Groß, Rolf Fritz}, title = {Interdisciplinary Course Smart Building Engineering: A new approach of teaching freshmen in remote teamwork project under pandemic restrictions}, series = {New Perspectives in Science Education - International Conference}, booktitle = {New Perspectives in Science Education - International Conference}, publisher = {Filodiritto}, address = {Bologna}, pages = {4 Seiten}, year = {2021}, abstract = {In the context of the Corona pandemic and its impact on teaching like digital lectures and exercises a new concept especially for freshmen in demanding courses of Smart Building Engineering became necessary. As there were hardly any face-to-face events at the university, the new teaching concept should enable a good start into engineering studies under pandemic conditions anyway and should also replace the written exam at the end. The students should become active themselves in small teams instead of listening passively to a lecture broadcast online with almost no personal contact. For this purpose, a role play was developed in which the freshmen had to work out a complete solution to the realistic problem of designing, construction planning and implementing a small guesthouse. Each student of the team had to take a certain role like architect, site manager, BIM-manager, electrician and the technitian for HVAC installations. Technical specifications must be complied with, as well as documentation, time planning and cost estimate. The final project folder had to contain technical documents like circuit diagrams for electrical components, circuit diagrams for water and heating, design calculations and components lists. On the other hand construction schedule, construction implementation plan, documentation of the construction progress and minutes of meetings between the various trades had to be submitted as well. In addition to the project folder, a model of the construction project must also be created either as a handmade model or as a digital 3D-model using Computer-aided design (CAD) software. The first steps in the field of Building information modelling (BIM) had also been taken by creating a digital model of the building showing the current planning status in real time as a digital twin. This project turned out to be an excellent training of important student competencies like teamwork, communication skills, and self -organisation and also increased motivation to work on complex technical questions. The aim of giving the student a first impression on the challenges and solutions in building projects with many different technical trades and their points of view was very well achieved and should be continued in the future.}, language = {en} } @inproceedings{SchulzeBuxlohGross2021, author = {Schulze-Buxloh, Lina and Groß, Rolf Fritz}, title = {Miniature urban farming plant: a complex educational "Toy" for engineering students}, series = {The Future of Education 11th Edition 2021}, booktitle = {The Future of Education 11th Edition 2021}, pages = {4 Seiten}, year = {2021}, abstract = {Urban farming is an innovative and sustainable way of food production and is becoming more and more important in smart city and quarter concepts. It also enables the production of certain foods in places where they usually dare not produced, such as production of fish or shrimps in large cities far away from the coast. Unfortunately, it is not always possible to show students such concepts and systems in real life as part of courses: visits of such industry plants are sometimes not possible because of distance or are permitted by the operator for hygienic reasons. In order to give the students the opportunity of getting into contact with such an urban farming system and its complex operation, an industrial urban farming plant was set up on a significantly smaller scale. Therefore, all needed technical components like water aeriation, biological and mechanical filtration or water circulation have been replaced either by aquarium components or by self-designed parts also using a 3D-printer. Students from different courses like mechanical engineering, smart building engineering, biology, electrical engineering, automation technology and civil engineering were involved in this project. This "miniature industrial plant" was also able to start operation and has now been running for two years successfully. Due to Corona pandemic, home office and remote online lectures, the automation of this miniature plant should be brought to a higher level in future for providing a good control over the system and water quality remotely. The aim of giving the student a chance to get to know the operation of an urban farming plant was very well achieved and the students had lots of fun in "playing" and learning with it in a realistic way.}, language = {en} } @inproceedings{SchulzeBuxlohGrossUlbrich2021, author = {Schulze-Buxloh, Lina and Groß, Rolf Fritz and Ulbrich, Michelle}, title = {Digital planning using building information modelling and virtual reality: new approach for students' remote practical training under lockdown conditions in the course of smart building engineering}, series = {Proceedings of International Conference on Education in Mathematics, Science and Technology 2021}, booktitle = {Proceedings of International Conference on Education in Mathematics, Science and Technology 2021}, publisher = {ISTES Organization}, address = {San Antonio, TX}, isbn = {978-1-952092-17-6}, pages = {118 -- 123}, year = {2021}, abstract = {The worldwide Corona pandemic has severely restricted student projects in the higher semesters of engineering courses. In order not to delay the graduation, a new concept had to be developed for projects under lockdown conditions. Therefore, unused rooms at the university should be digitally recorded in order to develop a new usage concept as laboratory rooms. An inventory of the actual state of the rooms was done first by taking photos and listing up all flaws and peculiarities. After that, a digital site measuring was done with a 360° laser scanner and these recorded scans were linked to a coherent point cloud and transferred to a software for planning technical building services and supporting Building Information Modelling (BIM). In order to better illustrate the difference between the actual and target state, two virtual reality models were created for realistic demonstration. During the project, the students had to go through the entire digital planning phases. Technical specifications had to be complied with, as well as documentation, time planning and cost estimate. This project turned out to be an excellent alternative to on-site practical training under lockdown conditions and increased the students' motivation to deal with complex technical questions.}, language = {en} } @inproceedings{BedburMoorkampPetersonetal.2020, author = {Bedbur, Christian and Moorkamp, Wilfried and Peterson, Leif Arne and Uibel, Thomas}, title = {Holzbr{\"u}cken mit Natursteinbel{\"a}gen - Untersuchungen zur Aktivierung der Verbundwirkung}, series = {Doktorandenkolloquium Holzbau Forschung + Praxis 2020}, booktitle = {Doktorandenkolloquium Holzbau Forschung + Praxis 2020}, editor = {Kuhlmann, Ulrike}, publisher = {Universit{\"a}t Stuttgart, Institut f{\"u}r Konstruktion und Entwurf}, address = {Stuttgart}, pages = {53 -- 60}, year = {2020}, language = {de} } @inproceedings{IomdinaKiselevaKotliaretal.2020, author = {Iomdina, Elena N. and Kiseleva, Anna A. and Kotliar, Konstantin and Luzhnov, Petr V.}, title = {Quantification of Choroidal Blood Flow Using the OCT-A System Based on Voxel Scan Processing}, series = {Proceedings of the International Conference on Biomedical Innovations and Applications- BIA 2020}, booktitle = {Proceedings of the International Conference on Biomedical Innovations and Applications- BIA 2020}, publisher = {IEEE}, address = {New York, NY}, isbn = {978-1-7281-7073-2}, doi = {10.1109/BIA50171.2020.9244511}, pages = {41 -- 44}, year = {2020}, abstract = {The paper presents a method for the quantitative assessment of choroidal blood flow using an OCT-A system. The developed technique for processing of OCT-A scans is divided into two stages. At the first stage, the identification of the boundaries in the selected portion was performed. At the second stage, each pixel mark on the selected layer was represented as a volume unit, a voxel, which characterizes the region of moving blood. Three geometric shapes were considered to represent the voxel. On the example of one OCT-A scan, this work presents a quantitative assessment of the blood flow index. A possible modification of two-stage algorithm based on voxel scan processing is presented.}, language = {en} } @inproceedings{HauggKreyerKemperetal.2020, author = {Haugg, Albert Thomas and Kreyer, J{\"o}rg and Kemper, Hans and Hatesuer, Katerina and Esch, Thomas}, title = {Heat exchanger for ORC. adaptability and optimisation potentials}, series = {IIR International Rankine 2020 Conference}, booktitle = {IIR International Rankine 2020 Conference}, doi = {10.18462/iir.rankine.2020.1224}, pages = {10 Seiten}, year = {2020}, abstract = {The recovery of waste heat requires heat exchangers to extract it from a liquid or gaseous medium into another working medium, a refrigerant. In Organic Rankine Cycles (ORC) on Combustion Engines there are two major heat sources, the exhaust gas and the water/glycol fluid from the engine's cooling circuit. A heat exchanger design must be adapted to the different requirements and conditions resulting from the heat sources, fluids, system configurations, geometric restrictions, and etcetera. The Stacked Shell Cooler (SSC) is a new and very specific design of a plate heat exchanger, created by AKG, which allows with a maximum degree of freedom the optimization of heat exchange rate and the reduction of the related pressure drop. This optimization in heat exchanger design for ORC systems is even more important, because it reduces the energy consumption of the system and therefore maximizes the increase in overall efficiency of the engine.}, language = {en} } @inproceedings{AlexopoulosBreitbachHoffschmidt2009, author = {Alexopoulos, Spiros and Breitbach, Gerd and Hoffschmidt, Bernhard}, title = {Optimization of the channel form geometry of porous ReSiC ceramic membrane modules}, series = {Proceedings / International Conference \& Exhibition for Filtration and Separation Technology, FILTECH 2009 : October 13 - 15, 2009, Wiesbaden, Germany. Vol. 2}, booktitle = {Proceedings / International Conference \& Exhibition for Filtration and Separation Technology, FILTECH 2009 : October 13 - 15, 2009, Wiesbaden, Germany. Vol. 2}, publisher = {Filtech Exhibitions Germany}, address = {Meerbusch}, isbn = {978-3-941655-36-2}, pages = {686 -- 693}, year = {2009}, language = {en} } @inproceedings{IlleSchrammAdametal.2009, author = {Ille, Fabian and Schramm, S. and Adam, M. and Backes, K. and Anthrakidis, Anette and Faber, Christian}, title = {Kraft-W{\"a}rme-K{\"a}lte-Kopplung im Leistungsbereich in 10kW mit periodisch arbeitender Sorptionsmaschine}, series = {DKV-Tagungsbericht. 35,2,1. Deutsche K{\"a}lte-Klima-Tagung 2008 Ulm}, booktitle = {DKV-Tagungsbericht. 35,2,1. Deutsche K{\"a}lte-Klima-Tagung 2008 Ulm}, publisher = {Dt. K{\"a}lte- und Klimatechnischer Verein}, address = {Hannover}, isbn = {978-3-932715-44-0}, pages = {305 -- 320}, year = {2009}, language = {de} } @inproceedings{AhlbrinkAlexopoulosAnderssonetal.2009, author = {Ahlbrink, Nils and Alexopoulos, Spiros and Andersson, Joel A. E. and Belhomme, B. and Teixeira Boura, Cristiano Jos{\´e} and Gall, Jan and Hirsch, Tobias}, title = {viCERP - the Virtual Institute of Central Receiver Power Plant}, series = {MATHMOD 2009 - 6th Vienna International Conference on Mathematical Modelling : February 11 - 13, 2009, Vienna, Austria. ARGESIM Report. No. 35}, booktitle = {MATHMOD 2009 - 6th Vienna International Conference on Mathematical Modelling : February 11 - 13, 2009, Vienna, Austria. ARGESIM Report. No. 35}, publisher = {Elsevier}, address = {Amsterdam}, isbn = {978-3-901608-35-3}, year = {2009}, language = {en} } @inproceedings{FunkeBeckmannKeinzetal.2017, author = {Funke, Harald and Beckmann, Nils and Keinz, Jan and Abanteriba, Sylvester}, title = {Numerical and Experimental Evaluation of a Dual-Fuel Dry-Low-NOx Micromix Combustor for Industrial Gas Turbine Applications}, series = {Proceedings of the ASME Turbo Expo 2017: Turbomachinery Technical Conference and Exposition. Volume 4B: Combustion, Fuels and Emissions. Charlotte, North Carolina, USA. June 26-30, 2017}, booktitle = {Proceedings of the ASME Turbo Expo 2017: Turbomachinery Technical Conference and Exposition. Volume 4B: Combustion, Fuels and Emissions. Charlotte, North Carolina, USA. June 26-30, 2017}, publisher = {ASME}, address = {New York}, isbn = {978-0-7918-5085-5}, doi = {10.1115/GT2017-64795}, year = {2017}, abstract = {The Dry-Low-NOx (DLN) Micromix combustion technology has been developed originally as a low emission alternative for industrial gas turbine combustors fueled with hydrogen. Currently the ongoing research process targets flexible fuel operation with hydrogen and syngas fuel. The non-premixed combustion process features jet-in-crossflow-mixing of fuel and oxidizer and combustion through multiple miniaturized flames. The miniaturization of the flames leads to a significant reduction of NOx emissions due to the very short residence time of reactants in the flame. The paper presents the results of a numerical and experimental combustor test campaign. It is conducted as part of an integration study for a dual-fuel (H2 and H2/CO 90/10 Vol.\%) Micromix combustion chamber prototype for application under full scale, pressurized gas turbine conditions in the auxiliary power unit Honeywell Garrett GTCP 36-300. In the presented experimental studies, the integration-optimized dual-fuel Micromix combustor geometry is tested at atmospheric pressure over a range of gas turbine operating conditions with hydrogen and syngas fuel. The experimental investigations are supported by numerical combustion and flow simulations. For validation, the results of experimental exhaust gas analyses are applied. Despite the significantly differing fuel characteristics between pure hydrogen and hydrogen-rich syngas the evaluated dual-fuel Micromix prototype shows a significant low NOx performance and high combustion efficiency. The combustor features an increased energy density that benefits manufacturing complexity and costs.}, language = {en} } @inproceedings{ThomaStiemerBraunetal.2023, author = {Thoma, Andreas and Stiemer, Luc and Braun, Carsten and Fisher, Alex and Gardi, Alessandro G.}, title = {Potential of hybrid neural network local path planner for small UAV in urban environments}, series = {AIAA SCITECH 2023 Forum}, booktitle = {AIAA SCITECH 2023 Forum}, publisher = {AIAA}, address = {Reston, Va.}, doi = {10.2514/6.2023-2359}, pages = {13 Seiten}, year = {2023}, abstract = {This work proposes a hybrid algorithm combining an Artificial Neural Network (ANN) with a conventional local path planner to navigate UAVs efficiently in various unknown urban environments. The proposed method of a Hybrid Artificial Neural Network Avoidance System is called HANNAS. The ANN analyses a video stream and classifies the current environment. This information about the current Environment is used to set several control parameters of a conventional local path planner, the 3DVFH*. The local path planner then plans the path toward a specific goal point based on distance data from a depth camera. We trained and tested a state-of-the-art image segmentation algorithm, PP-LiteSeg. The proposed HANNAS method reaches a failure probability of 17\%, which is less than half the failure probability of the baseline and around half the failure probability of an improved, bio-inspired version of the 3DVFH*. The proposed HANNAS method does not show any disadvantages regarding flight time or flight distance.}, language = {en} } @inproceedings{BergmannGraebenerWildetal.2019, author = {Bergmann, Kevin and Gr{\"a}bener, Josefine and Wild, Dominik and Ulfers, H. and Czupalla, Markus}, title = {Study on thermal stabilization of a GEO-stationary telescope baffling system by integral application of phase change material}, series = {International Conference on Environmental Systems}, booktitle = {International Conference on Environmental Systems}, pages = {1 -- 14}, year = {2019}, abstract = {The utilization of phase change material (PCM) for latent heat storage and thermal control of spacecraft has been demonstrated in the past in few missions only. One limiting factor was the fact that all concepts developed so far envisioned the PCM to be applied as an additional capacitor, encapsulated in its own housing, leading to mass, efficiency and accommodation challenges. Recently, the application of PCM within the scan cavity of a GEOS type satellite has been suggested, in order to tackle thermal issues due to direct sun intrusion (Choi, M., 2014). However, the application of PCM in such complex mechanical structures is extremely challenging. A new concept to tackle this issue is currently under development at the FH Aachen University of Applied Sciences. The concept "Infused Thermal Solutions (ITS)" is based on the idea to 3D print metallic structures in their regular functional shape, but double walled with internal lattice support structures, allowing the infusion of a PCM layer directly into the voids and eliminating the need for additional parts and interfaces. Together with OHB System, FH Aachen theoretically studied the application of this technology to the Meteosat Third Generation (MTG) Infra-Red Sounder (IRS) instrument. The study focuses on the scan cavity and entrance baffling assembly (EBA) of the IRS. It consists of thermal analyses, 3D-redesign and bread boarding of a scaled and PCM infused EBA version. In the thermal design of the alternative EBA, PCM was applied directly into the EBA, simulating the worst hot case sun intrusion of the mission. By applying 4kg of PCM (to a 60kg baffle) the EBA temperature excursions during sun intrusion were limited from 140K to 30K, leading to a significant thermo-opto-elastic performance gain. This paper introduces the ITS concept development status.}, language = {en} } @inproceedings{StarkRiepingEsch2023, author = {Stark, Ralf and Rieping, Carla and Esch, Thomas}, title = {The impact of guide tubes on flow separation in rocket nozzles}, series = {Aerospace Europe Conference 2023 - 10th EUCASS - 9th CEAS}, booktitle = {Aerospace Europe Conference 2023 - 10th EUCASS - 9th CEAS}, pages = {8 Seiten}, year = {2023}, abstract = {Rocket engine test facilities and launch pads are typically equipped with a guide tube. Its purpose is to ensure the controlled and safe routing of the hot exhaust gases. In addition, the guide tube induces a suction that effects the nozzle flow, namely the flow separation during transient start-up and shut-down of the engine. A cold flow subscale nozzle in combination with a set of guide tubes was studied experimentally to determine the main influencing parameters.}, language = {en} } @inproceedings{StarkBartelDitscheetal.2023, author = {Stark, Ralf and Bartel, Sebastian and Ditsche, Florian and Esch, Thomas}, title = {Design study of a 30kN LOX/LCH4 aerospike rocket engine for lunar lander application}, series = {Aerospace Europe Conference 2023 - 10th EUCASS - 9th CEAS}, booktitle = {Aerospace Europe Conference 2023 - 10th EUCASS - 9th CEAS}, pages = {9 Seiten}, year = {2023}, abstract = {Based on lunar lander concept EL3, various LOX/CH4 aerospike engines were studied. A distinction was made between single and cluster configurations as well as ideal and non-ideal contour concepts. It could be shown that non-ideal aerospike engines promise a significant payload gain.}, language = {en} } @inproceedings{ElMoussaouiKassmiAlexopoulosetal.2021, author = {El Moussaoui, Noureddine and Kassmi, Khalil and Alexopoulos, Spiros and Schwarzer, Klemens and Chayeb, Hamid and Bachiri, Najib}, title = {Simulation studies on a new innovative design of a hybrid solar distiller MSDH alimented with a thermal and photovoltaic energy}, series = {Materialstoday: Proceedings}, volume = {45}, booktitle = {Materialstoday: Proceedings}, number = {8}, publisher = {Elsevier}, address = {Amsterdam}, issn = {2214-7853}, doi = {10.1016/j.matpr.2021.03.115}, pages = {7653 -- 7660}, year = {2021}, abstract = {In this paper, we present the structure, the simulation the operation of a multi-stage, hybrid solar desalination system (MSDH), powered by thermal and photovoltaic (PV) (MSDH) energy. The MSDH system consists of a lower basin, eight horizontal stages, a field of four flat thermal collectors with a total area of 8.4 m2, 3 Kw PV panels and solar batteries. During the day the system is heated by thermal energy, and at night by heating resistors, powered by solar batteries. These batteries are charged by the photovoltaic panels during the day. More specifically, during the day and at night, we analyse the temperature of the stages and the production of distilled water according to the solar irradiation intensity and the electric heating power, supplied by the solar batteries. The simulations were carried out in the meteorological conditions of the winter month (February 2020), presenting intensities of irradiance and ambient temperature reaching 824 W/m2 and 23 °C respectively. The results obtained show that during the day the system is heated by the thermal collectors, the temperature of the stages and the quantity of water produced reach 80 °C and 30 Kg respectively. At night, from 6p.m. the system is heated by the electric energy stored in the batteries, the temperature of the stages and the quantity of water produced reach respectively 90 °C and 104 Kg for an electric heating power of 2 Kw. Moreover, when the electric power varies from 1 Kw to 3 Kw the quantity of water produced varies from 92 Kg to 134 Kg. The analysis of these results and their comparison with conventional solar thermal desalination systems shows a clear improvement both in the heating of the stages, by 10\%, and in the quantity of water produced by a factor of 3.}, language = {en} } @inproceedings{WalterElsenMuelleretal.1999, author = {Walter, Peter and Elsen, Ingo and M{\"u}ller, Holger and Kraiss, Karl-Friedrich}, title = {3D object recognition with a specialized mixtures of experts architecture}, series = {IJCNN'99. International Joint Conference on Neural Networks. Proceedings}, booktitle = {IJCNN'99. International Joint Conference on Neural Networks. Proceedings}, publisher = {IEEE}, address = {New York}, isbn = {0-7803-5529-6}, issn = {1098-7576}, doi = {10.1109/IJCNN.1999.836243}, pages = {3563 -- 3568}, year = {1999}, abstract = {Aim of the AXON2 project (Adaptive Expert System for Object Recogniton using Neuml Networks) is the development of an object recognition system (ORS) capable of recognizing isolated 3d objects from arbitrary views. Commonly, classification is based on a single feature extracted from the original image. Here we present an architecture adapted from the Mixtures of Eaqerts algorithm which uses multiple neuml networks to integmte different features. During tmining each neural network specializes in a subset of objects or object views appropriate to the properties of the corresponding feature space. In recognition mode the system dynamically chooses the most relevant features and combines them with maximum eficiency. The remaining less relevant features arz not computed and do therefore not decelerate the-recognition process. Thus, the algorithm is well suited for ml-time applications.}, language = {en} }