Scientific direction Development of key enabling technologies
Transfer of knowledge to industry

PhD : selection by topics

Incorporating expert knowledge and linguistic resources in deep neural networks for multi-domain and multilingual adaptation

Département Intelligence Ambiante et Systèmes Interactifs (LIST)

Vision & Ingénierie des Contenus (SAC)

01-03-2019

SL-DRT-19-0750

nasredine.semmar@cea.fr

Recent deep learning architectures and algorithms have shown impressive results for several Natural Language Processing (NLP) tasks such as Named entity recognition, Part-of-Speech tagging, Dependency parsing and Semantic role labelling. The actual performance of certain NLP tools for English evaluated on in-domain data is close to human level, thanks to deep learning models trained on huge annotated datasets. Contrariwise, approaching human-level accuracy on more complex domains and low-resource languages is still a hard issue. The proposed subject aims to explore and experiment new approaches based on the integration of expert knowledge and available linguistic resources in deep neural networks in order to improve the performance of NLP tools for specialty areas and low-resource languages. We propose to tackle this issue along the following key aspects, as an extension of the research work already carried out at LVIC laboratory: - Taking into account heterogeneous expert knowledge and linguistic resources: Ontologies, Terminology databases, Lexicons, Named entity recognition rules, Dependency parsing recognition rules, etc. - Implementing a formalism to describe expert knowledge and linguistic resources in a multi-level representation. The objective is to define a structure in which the different types of expert knowledge and linguistic resources will be represented separately but the whole representation would be described in the same format (model). - Exploring new strategies for incorporating expert knowledge and linguistic resources in deep neural networks. The underlying idea is to propose an integration mechanism which can be adapted to each expert knowledge and linguistic resource.

Development of metal components by stereolithography - Copper for energy application

Département des Technologies des NanoMatériaux (LITEN)

Laboratoire Recyclage et Valorisation des Matériaux

01-09-2019

SL-DRT-19-0751

marilyne.roumanie@cea.fr

Stereolithography (SLA) is a well-known additive manufacturing process based on photopolymerisation reaction. Specifically, metal SLA is based on the blend of a photocurable resin with a metallic powder. In this process, the polymer gives mechanical strength to the part during the manufacturing step. The polymer is then thermally removed (debinding step), and the parts is sintered. This process has the benefit of addressing a large number of materials, relying on the expertise of the powder metallurgy in terms of debinding /sintering and being easily implemented in this industrial field. Copper is known for energy application by their high thermal (385 W/m.K) and electrical (59.6 x 106 S.m-1) properties. The objective is to control the impact of the stereolithography process (formulation, light/matter interaction, thermal treatment) on thermal and electrical properties of bulk and structured copper components. The impact of the formulation on the cross-linking rate and the green mechanical strength, the relation between the cross-linking rate and the filled resin degradation on the cracking during the thermal step, and the relation between the thermal and electrical properties of the copper components and the microstructure, the density and the light elements contamination should be investigated during the PhD.

MILP models for optimal management of hybrid CSP plants

Département Thermique Biomasse et Hydrogène (LITEN)

Laboratoire Systèmes Solaires Haute Température

01-12-2019

SL-DRT-19-0769

valery.vuillerme@cea.fr

The hybridisation of CSP plants with "conventional" plants (existing or not) has many advantages, but in return brings more complexity in the control of the system and the steering strategy. At present, we do not find in the study literature dealing with this issue and highlighting the principles of predictive management of hybrid CSP plants. The planned work will allow the development of dynamic models of CSP plants by taking advantage of the special possibilities offered by the Cathare code in co-simulation environment (PEGASE). This environment will address the advanced control-command aspects of hybrid CSP systems (biomass, geothermal, incineration, coal, nuclear, H2 ...), and we will implement examples to demonstrate the contribution of predictive management to such systems constrained by variations in demand (market) and resource. For validation purposes, certain internal and external experimental means will be used, such as the prototypes of the solar zone at Cadarache for elements relating to CSP / storage pairs, or infrastructures made available under the SFERA III project to which participates the CEA. Ultimately, the numerical demonstrator developed will highlight the hybrid CSP systems of major interest at national, European and international level.

Numerical and experimental studies of a combined cooling and power cycle

Département Thermique Biomasse et Hydrogène (LITEN)

Laboratoire Systèmes Solaires Haute Température

01-10-2019

SL-DRT-19-0770

haitrieu.phan@cea.fr

The TRICYCLE project aims to study and develop a thermodynamic heat recovery cycle for the combined production of cold and low power electricity (5 kW of cold, 1 kW of electricity). The target temperature range is sufficiently low (80 to 160 °C) to target industrial heat recovery applications, but also heat networks and non-concentrated solar thermal energy. In the previous work, the thermodynamic modeling of the cycle was carried out and allowed the definition of an parallel architecture: the refrigerant vapor at the output of the generator can be used to feed the expander(production of electricity) and/or the condensation-expansion-evaporation part (production of cold). This is a unique cycle and not a juxtaposition of two machines (cold + generator), which makes it a particularly innovative system. The objective in 2019 of the project is to achieve an operational prototype of this combined cycle by integrating an expander to the existing absorption-chiller available in our laboratory. This thesis is the next steps of the project in order to: ? Conduct the experimental measurements of the TRICYCLE machine using an adapted instrumentation in order to better understand its behavior (static and dynamic) ? Conduct numerical simulations of the machine (under DYMOLA / MODELICA) and validate the models by comparison to the measurements ? Simulate the coupling of this machine with add-ons (heat storage and / or electrical storage) in a system (application: heat network, automobile, etc.).

Experimental and Theoretical Study of the Scaling Limits of GaN Power Electronics Devices for High Efficiency Converters

Département Composants Silicium (LETI)

Laboratoire Composants Electroniques pour l'Energie

01-09-2019

SL-DRT-19-0783

julien.buckley@cea.fr

Background: High Electron Mobility Transistors (HEMT) using a heterostructure built on gallium nitride (GaN) are highly promising in the field of power electronics. GaN is a wide bandgap semiconductor that can withstand high electric fields while allowing low conduction losses due to the high mobility of the carriers in the 2 Dimensional Electron Gas (2DEG) at the heterostructure interface. Proposed work: The study will focus on the identification of the size limitations of current devices by using the electrical characterization results of componants with multiple geometries, already fabricated at CEA. Finite element simulations (with SYNOPSYS software) will be performed in order to interpret the electrical results and test multiple hypotheses regarding their physical operation. Expected results: The advances achieved regarding the understanding of the devices will be used in order to identify improvements of the architecture studied by simulation. An electrical evaluation of the proposed solutions will be performed after fabrication of the new components in collaboration with the device integration and process development teams.

Wet Electrostatic Precipitation to reduce urban air pollution

Département Microtechnologies pour la Biologie et la Santé (LETI)

Laboratoire Biologie et Architecture Microfluidiques

01-10-2019

SL-DRT-19-0793

jean-maxime.roux@cea.fr

Air pollution, especially urban air pollution, is a public health problem leading in France to nearly 50 000 deaths per year. In 2018 the World Health Organization reported that toxic levels of pollution leads annually to the early death of about 7 million people. The PhD subject deals with the design of a new urban clean-up system based on wet electrostatic precipitation. Air purifiers based on this principle are usually intended for an industrial use. The PhD will focus on a multiphysical numerical simulation of such a device, but adapted to an urban deployment, starting with the central and difficult problem posed by the stability of an air/water interface in an intense electric field. While being based on this simulation, the final challenge is to develop a numerical optimization of the system aiming at a significant reduction of its size and an appropriate integration of the toxic gas / airborne particles sensors developed at CEA GRENOBLE/Leti/DTBS. Experimental studies carried out at CEA will be guided by the obtained numerical results which will in return be validated.

108 Results found (Page 12 of 18)
first   previous  10 - 11 - 12 - 13 - 14  next   last

Voir toutes nos offres