Université de technologie de Troyes open archive
Not a member yet
10722 research outputs found
Sort by
Strategic Energy Management in Multi-Source Microgrids: A Bertrand Duopoly Game Algorithm Model
No full textInternational audienceThis paper presents an innovative energy management strategy for a grid-connected multi-source system, utilizing a Bertrand duopoly game model. This approach integrates a pricing mechanism and energy availability togovern the interactions between suppliers and the consumer. We propose a negotiation-based strategy thatsimulates these interactions, with all suppliers participating on equal terms by simultaneously setting theirselling prices. In response, to the varying prices, the consumer selects the most favorable energy source. Thisstrategy balances energy distribution, minimizing costs for the consumer while maximizing energy utilization forthe suppliers. The performance and effectiveness of the proposed algorithm were validated through c extensiveMATLAB/Simulink simulations. The results indicate that the model enhances operational efficiency and promotes sustainable energy management practices within hybrid energy systems. The integration of game theoryallows for the optimization of interactions among stakeholders and addresses the complexities of decisionmaking in multi-source systems
Two-dimensional cutting stock problem with flexible length and usable leftovers in the steel industry
No full textInternational audienceIn this work, we introduce a two-dimensional cutting stock problem with flexible length and usable leftovers, in which multiple objectives, including minimizing the waste area of material, the exceeding area of orders and the number of slitter adjustments, are considered simultaneously. This problem is inspired by a real-world made-to-order manufacturer of special steel plates. We propose a non-linear mathematical programming model for this problem. This model is then linearized and reinforced by symmetry-breaking inequalities and other valid inequalities. To solve this model, we propose an iterated local search algorithm, which is able to tackle large instances of the problem. Numerical results demonstrate the validity of the proposed model and the effectiveness of the iterated local search algorithm. Besides, sensitive experiments are conducted to assess the impact of different parameter settings on the objective function
Managing no-shows and resource utilisation for outpatient appointment scheduling
No full textInternational audienceOutpatient appointment scheduling is complicated by uncertainties such as patient no-shows and variability in service times, leading to under-utilisation of resources and patient dissatisfaction. While overbooking strategies mitigate the impact of no-shows, achieving an optimal balance between resource optimisation and patient satisfaction remains challenging. In this study, we propose a novel approach to outpatient appointment scheduling by applying the newsvendor problem framework to manage overbooking and reduce the negative effects of no-shows. Our model optimizes the overbooking threshold to maximize expected profit while balancing resource utilisation and patient satisfaction, considering service time variability. A sensitivity analysis highlights how key parameters influence the optimal overbooking level, providing a quantitative solution to balance patient waiting times and physician idle times. Our findings offer actionable insights for healthcare providers seeking to improve clinic efficiency
A hybrid metaheuristic to solve the multi-trip team orienteering problem with integrated tours and increasing profits for blood transportation
No full textInternational audienceThe efficient management of the blood supply chain is paramount, given its critical impact on human lives if improperly handled. This research focuses on the location of mobile collection sites over a multiperiod horizon and transferring blood units from these sites to blood centers using a multi-trip shuttle fleet. The goal is to minimize the total cost comprising the transportation, delay, wastage, and shortage costs. The model considers constraints such as the tight time limit between the collection and processing of platelets and cryoprecipitates. Additionally, collection levels at donation sites are modeled assuming linear increasing rates. First, the problem is modeled as a mixed-integer linear program; then, a hybrid iterative local search is designed as a solution approach. The effectiveness and applicability of the proposed method are evaluated through a set of 39 new instances based on the blood collection system of Bogota, Colombia, which demonstrates promising solutions by systematically computing the best-known solutions
Characterizing the ingenuity of the Low-Tech design approach: a driver to move design toward more sustainable practices
No full textInternational audienceIn contexts of low resources and under the specter of socio-ecological challenges, various alternative design approaches have emerged such as Jugaad, Gambiarra, Low-Tech, etc. They are characterized by the ingenuity shown by designers in creating solutions under constraints. However, the capacity for ingenuity in this context has not yet been studied and this work proposes a first step to characterize it. The focus is on one design approach: The Low-Tech approach. Two hypotheses are proposed to characterize the ingenuity of Low-Tech designers. First, unlike conventional design contexts, Low-Tech designers use self-imposed ethical constraints during the design process. Second, in the way they operate, Low-Tech designers are closer to the logic of the « bricoleur » than to the logic of the engineer. A mixed method approach was used, involving quantitative data to identify the relevant phenomenon (large-scale questionnaire), followed by qualitative data to better grasp its complexity (semi-structured interviews with 10 designers). This research provides a better understanding of the ingenuity demonstrated by Low-Tech designers. Ultimately, the aim is to pass on this ability to conventional designers in the hope that it will lead to a paradigm shift toward a more sustainable one
Disposable biosensors based on nanostructured substrates for effective monitoring of human biomarkers and environmental contaminants (Keynote)
No full textInternational audienceMore than ever, the world is facing the risk of pandemics, which can significantly affect thenormal functioning of hospitals, with a direct impact on public health. As a result, many patients may be forced to interrupt or postpone their treatment/surgery due to excessive demand for emergency services and exceptional pressure on hospital facilities. Therefore, there is an urgent need for rapid and sensitive methods for identifying/assessing in real-time the main types of biomarkers in body fluids. In this context, biosensors are ideal candidates [1], as they can be designed as single-use devices for research/medical laboratories and general public interested in assessing specific biomarkers in various biological samples (such as blood, urine, saliva, etc.) at home. Therefore, the presence of biological and environmental markers is monitored by measuring/comparing various parameters: color (colorimetric) [2], light peak intensity (nanoplasmonic) [3], current (electrochemical) [2], or frequency (acoustic) [4] as a result of interaction with biofunctionalized substrates with specific antibodies, antigens, enzymes, orcomplementary DNA sequence. It should be emphasized that one of the most importantparameters for achieving a lower limit of detection (LOD) for a particular (biological) markeris the choice of substrate, which can be solid (e.g. glass) or flexible (e.g. absorbent cellulosepaper), fully or partially tailored with nanomaterials (e.g. gold nanoparticles) [5]. Subsequently, the collected bodily fluids are applied drop by drop directly onto such accessible and inexpensive substrates developed for a specific type of biosensor. The presentation will focus on the development of different biosensor configurations using nanostructured substrates, highlighting the particular importance of optimal experimental conditions (substrate selection, running biological buffers, incubation times, choice of controls, etc.) to significantly improve the analytical performances. The perspectives of the next generation of multi-transducer biosensors for the detection of (bio)markers will also be discussed
Pulsed laser sources for nanometer-scaled complex materials and devices
No full textInternational audienceThis communication reports on the development and synthesis of nanometer-scale complex architectures, including nanoparticles and thin films, fabricated using Pulsed Laser Deposition (PLD) combined with a free cluster generator. The association of the two technics results in the elaboration of various materials with a wide range of applications.PLD facilitates the deposition of films with nanometer-scale thicknesses control, high density and crystallinity, ensuring congruent transfer from the target to the film. The synthesis of nanoparticles (NPs) is achieved using a free cluster generator developed by our group, based on Smalley’s source principle. The pulsed plasma plume generated by a laser focused on a target is quenched in a nucleation chamber by a synchronized helium puff delivered through a controlled pulsed valve. This quenching process allows the formation of nuclei and the subsequent cluster growth. The NPs passed inside the PLD chamber through a nozzle designed to maintain their size and shape from the nucleation chamber. Using separately, simultaneously or sequentially the NPs generator and/or the PLD, leads to the synthesis of new nanomaterials with exotic properties, respectively NP-stacks, nanocomposite thin films with embedded NPs or NPs/thin film multilayers [1, 2].This specific setup was used to synthetize complex structures including Au, Al2O3 thin films and Ag NPs. Monodisperse and crystallized metallic silver NPs (diameter 2.5 ± 0.5 nm) incorporated into the amorphous Al2O3 matrix exhibits optical absorption at a wavelength of 432 nm, consistent with Mie theory. This reactor provides contamination-free interfaces, with an accurate control over each layer. The growth at ambient temperature additionally avoids interdiffusion between the different material components.These nanomaterials have a large set of applications particularly for electronics and optoelectronics, such as plasmonic based biosensors or solar cells, optical switching and modulation, transparent conductive films, microelectronic components as Metal-Insulator-Metal (MIM) capacitors, etc…This work was partially supported by the LabEx SigmaLim ANR-10-LBX-0074-01, laboratory of excellence launched by the French Ministry of Higher Education and Research between the XLIM and IRCER Research Institutes.[1] M. Gaudin et al., A dual nanosecond-pulsed laser setup for nanocomposite synthesis—Ag nanoparticles in Al2O3/VO2 matrix, J. Appl. Phys., vol. 125, no 5, p. 054301, 2019.[2] F. Dumas-Bouchiat et al., VO2 thin films: various microstructures for hysteresis manipulations, Vacuum, vol. 227, p. 113408, 2024
Degradation model selection using depth function: A comparative analysis of median and outlier of functional data
No full textInternational audienc
La place des grossistes dans la transition alimentaire en territoires : le cas du Marché d’Intérêt National de Strasbourg
No full textMasterAlors que le système agro-industriel dominant est progressivement remis en question du fait des externalités négatives qu’il génère, la résilience alimentaire est progressivement interrogée à l’échelle des territoires. Cette démarche s’institutionnalise à travers les projets alimentaires territoriaux (PAT). Ces derniers mobilisent peu les acteurs intermédiaires des systèmes alimentaires. Les grossistes implantés sur un Marché d’Intérêt National (MIN) occupent un rôle singulier, à l’interface d’intérêts publics portés par les collectivités territoriales et de leurs intérêts privés et marchands. Le MIN constitue une place de marché intermédiaire, éclairante de la façon dont les grossistes contribuent (ou pas) aux initiatives de territorialisation et de circularité des systèmes alimentaires. Ce mémoire étudie l’ancrage territorial du MIN de Strasbourg. L’étude des flux agri-alimentaire s’accompagne de l’analyse des logiques d’action des grossistes, pour décrire le fonctionnement dominé par la dimension économique, et dont les acteurs sont principalement mus par une logique marchande, dans un contexte concurrentiel fort. Les flux générés révèlent une majorité de produits alimentaires importés d’Europe et d’Afrique du Nord, et une zone de chalandise limitée essentiellement à la région Alsace. Ces flux et logiques d’actions sont comparés aux enjeux de territorialisation et de durabilité du PAT l’Eurométropole de Strasbourg, afin d’en extraire les points de convergence. Les modalités de coopérations avec des producteurs locaux, fondées sur un attachement commun au territoire, structurent une rationalisation économique et une optimisation logistique. Finalement, cette étude propose d’identifier les leviers et les freins à la transition alimentaire en croisant les enjeux communs aux grossistes et aux collectivités territoriales. Elle pose les bases d’une réflexion commune pour une meilleure interconnaissance de acteurs et la prise en compte de leurs spécificités respectives
RgeoJSD : une fonction de coût robuste au bruit d'annotation dérivée de la JSD géométrique pour la classification d'emboles cérébraux
No full textInternational audienceCerebral emboli, solid or gaseous particles circulating in the cerebral blood stream, can obstruct arteries and are a major cause of stroke. Transcranial Doppler (TCD) ultrasonography uniquely enables continuous, long-term monitoring, particularly with portable devices. Deep learning methods offer the potential to capture complex microemboli patterns in TCD signals, improving emboli identification. However, the development of such techniques relies heavily on labeled data, which are often noisy due to experts' uncertainty. Enhancing the robustness of training to label noise not only alleviates the experts' labeling burden but also facilitates the integration of unannotated data through semi-automatic annotation. In this work, we propose the Robust geometric Jensen-Shannon Divergence (RgeoJSD) loss function, which leverages the favorable mathematical properties of geometric JSD (geoJSD). We show that geoJSD can be decomposed into a cross-entropy-like term and an additional term that enhances noise robustness. By weighting these components, RgeoJSD achieves tolerance to label noise. We evaluated RgeoJSD on a TCD dataset comprising 1 232 labeled spectrograms from 35 subjects, recorded using a TCD-X Holter device. Spectrograms were annotated into three classes: solid emboli, gaseous emboli, and artifacts. Under synthetic symmetric noise conditions, RgeoJSD performed comparably to standard cross-entropy loss when labels were clean, and significantly outperformed it under noise, improving the accuracy by up to 4.4% and 9.9% at noise rates of 0.2 and 0.4, respectively. These findings indicate that RgeoJSD is a promising solution for training deep learning models on noisy TCD data, and a strong candidate for enabling reliable, automated cerebral emboli classification