Portail HAL de l'Université Picardie Jules Verne
Not a member yet
32425 research outputs found
Sort by
A review of multimodal large language models for smart grids management and control
No full textInternational audienceThe worldwide effort to reach carbon peak and neutrality objectives alongside energy market expansion has sped up renewable energy integration, like wind and solar power. The shift towards renewable energy integration introduces substantial uncertainties in power system scheduling and control processes, which test the limits of existing theoretical methods. The advanced reasoning and data-processing capabilities of Large Language Models (LLMs), with particular reference to their ability to analyze multimodal data, provide transformative potential for managing and controlling smart grids. This review examines how LLMs can tackle modern power system challenges while confirming their fit with the power sector’s expanding dependency on Artificial Intelligence (AI) technologies. We assess the requirements of modern power systems for such AI-based solutions, while evaluating how LLMs shape grid management and exploring their enabling technologies, such as model architecture and training methods, along with necessary data. Our review investigates how multimodal LLM technology serves different smart grids’ functions, including generation, transmission, distribution, consumption, and equipment management, to exhibit its adaptable nature in strengthening grid resilience and efficiency
Characterisation of the Novel HLA‐C*03:741 Allele by Next Generation Sequencing
No full textInternational audienceHLA‐C*03:741 differs from the HLA‐C*03:04:01:01 allele by one non‐synonymous nucleotide at position 729 in exon 2
A Hybrid Framework for Intelligent Crisis Management in Parkinson's Disease
No full textInternational audienceParkinson’s disease (PD) is a chronic neurodegenerative disorder with progressive motor symptoms(tremor, rigidity, gait freezing, falls) and disabling non-motor features such as depression,anxiety, sleep problems. Patients may suddenly deteriorate (severe OFF episodes, unexpectedfalls...) and require rapid medical intervention, often outside hospital settings [1]. This createsa strong need for timely detection of crises and efficient dispatch of appropriate emergencyresources. Existing research on EMS in PD makes little use of longitudinal clinical informationand digital biomarkers, rely on triage decisions based on incomplete data, and do not fullyexploit modern connectivity (Internet of medical things, wearables, V2X) [2]. In this work,we propose an Intelligent Dispatch framework that connects smartphone-based monitoring,machine learning (ML), eXplainable AI (XAI) and optimization within the context of a crisisrelated to Parkinson’s disease. The framework aims to detect clinically relevant deteriorationin real time, classify patients into urgency levels, provide interpretable explanations ofrisk, and use these predictions to guide ambulance and hospital selection, ultimately reducingtime-to-intervention and improving coordination of emergency care for people with PD
La simulation comme outil méthodologique pour améliorer la collaboration interprofessionnelle
No full textInternational audienc
Indigénat des espèces arborées de France à l’échelle des sylvoécorégions
No full textInternational audienceIndigénat des espèces arborées de France à l'échelle des sylvoécorégions RÉSUMÉ.Cet article présente la première caractérisation de l'indigénat des espèces arborées de France hexagonale à l'échelle des sylvoécorégions. Les sylvoécorégions constituent un découpage du territoire particulièrement adapté aux études forestières, basé sur des critères pédoclimatiques. Les taxons traités sont les espèces arborées au sens de l'Inventaire forestier national, c'est-à-dire les espèces ligneuses pouvant atteindre 5 mètres de haut à maturité. Mené par l'Institut national de l'information géographique et forestière, ce travail a été réalisé avec l'expertise des conservatoires botaniques nationaux et de spécialistes des questions d'indigénat. À l'issue de ce travail, 119 espèces sont considérées comme étant indigènes ou cryptogènes dans au moins une sylvoécorégion, et donc indigènes ou cryptogènes en France. Cette nouvelle information est maintenant disponible sur la plateforme de l'État « recherche. data. gouv » à l'adresse https://doi.org/10.57745/DHJHGS, et peut être utilisée dans les études nécessitant une approche fine de l'indigénat des espèces arborée
L’effondrement de la biodiversité floristique : résultat d’une rupture dans le mode d’exploitation des milieux semi-naturels
No full textInternational audienceFor centuries, the way farming communities exploited previously deforested lands have been perpetuated. Semi-natural environments established gradually, remarkably rich in flora and fauna. The profound transformations that affected farming practices caused a rupture that disrupted an ancestral balance. Concerning Western Picardy, this this collapse is easily perceptible in various environments. The abandonment of plots deemed unproductive concerns both dry calcareous grasslands and floodplain meadows. In cereal fields, the (a long time immoderate) use of fertilizers and herbicides has led to the near disappearance of commensal vegetation. The spread of urbanization, both in urban areas and on the outskirts of villages, has disrupted sites that had long been surveyed by local botanists. Resilience that could intervene to compensate for the damage caused is considered.L’effondrement de la biodiversité floristique est le résultat d’une rupture dans le mode d’exploitation des milieux semi-naturels. Pendant des siècles, les modalités de l’exploitation par les communautés paysannes des terrains préalablement déboisés se sont perpétuées. Progressivement ont été mis en place des milieux semi-naturels dont la richesse floristique et phytocénotique était remarquable. Les profondes transformations ayant affecté les pratiques agricoles ont provoqué une rupture qui a bouleversé un équilibre ancestral. En ce qui concerne la Picardie occidentale, il est possible de prendre en considération plusieurs milieux dans lesquels cet effondrement est aisément perceptible. L’abandon des parcelles jugées insuffisamment productives concerne aussi bien les pelouses crayeuses sèches que les milieux prairiaux inondables. De même, l’usage trop longtemps immodéré des herbicides et des fertilisants a entrainé une quasi-disparition de la végétation commensale dans les champs de céréales. L’extension de l’urbanisation particulièrement en milieu urbain comme à la périphérie des villages a bouleversé des sites prospectés de longue date par les botanistes locaux. Il importe de tirer parti de la notion de résilience de façon à intervenir afin de compenser les dégâts ayant été commis
Parler ou se taire ? L'injustice émotionnelle dans la recherche sur les VSS
No full textInternational audienc
Prognosis of pediatric restrictive cardiomyopathy: more severe in sarcomeric variants
No full textInternational audienceRestrictive cardiomyopathy (RCM) is the most severe type of cardiomyopathy in children with a very poor prognosis. RCM is often diagnosed between 6 and 10 years old and is predominantly of genetic origin. We conducted a retrospective study of 53 patients. The aim of our study was to determine whether outcomes are associated with the type of genetic variant. We compared the prognosis of patients with sarcomeric variants (n = 26) to those with non sarcomeric variants (n = 27). Our results showed no significant differences between the two groups at diagnosis in terms of heart failure symptoms, NT-proBNP levels, or hemodynamic parameters. However, survival without transplantation was significantly worse in the sarcomeric group (p = 0.003), which also exhibited greater disease severity. Furthermore, thrombotic events were more frequent in the sarcomeric group (p = 0.05). In conclusion, RCM caused by sarcomeric variants is associated with a poorer prognosis and a higher incidence of thrombotic events compared to non-sarcomeric RCM
