HAL Portal UBS (Université Bretagne Sud)
Not a member yet
14201 research outputs found
Sort by
Bilan des connaissances de la dynamique de l’érosion du littoral sénégalais
No full textInternational audienc
Reinforcement learning-based energy management for hybrid electric vehicles: A comprehensive up-to-date review on methods, challenges, and research gaps
No full textInternational audienceReinforcement learning is widely used for control applications and has also been successfully implemented for efficient energy management within hybrid electric vehicles. Reinforcement learning algorithms offer various advantages, including fast convergence, broad applicability, stability, and robustness, particularly with the integration of deep and transfer learning. This paper provides a comprehensive understanding of reinforcement learning principles and a critical review of various reinforcement learning methods, states, actions, and rewards used to optimize the energy management performance of hybrid electric vehicles. Furthermore, the advantages and limitations of these algorithms are also discussed. This review reveals that deep reinforcement learning techniques show superior performance in handling complex energy management tasks thanks to their ability to learn from high-dimensional state spaces. Nevertheless, their implementation faces notable obstacles, including computational complexity and generalization across diverse driving conditions. Finally, key research directions for future work and challenges are highlighted in the domain of reinforcement-learning-based hybrid electric vehicle energy management
Field and laboratory characterization of the albedo of asphalt and horizontal surfaces
No full textInternational audienceAbstract The rise in temperature in urban areas compared to surrounding rural zones is a major concern for modern cities. Part of this phenomenon is attributed to the widespread use of modern construction materials with low albedo, particularly for road pavements. While outdoor albedo measurement standards exist, there is currently no consensus on a standardized laboratory method. This article aims to develop a multi-material, laboratory-to-field scale law for predicting outdoor albedo using only laboratory tests. This predictive law is intended to facilitate the design of new light-colored pavements and surfaces, without the need for full-scale construction and testing. The American standard ASTM E1918 provides guidance for accurate and reliable albedo measurements on low-sloped surfaces; however, it presents several limitations, which are discussed in this paper, and it is demonstrated that albedo measurements can be reliably performed on-site under a broader range of conditions than those specified by the standard. Additionally, although some studies have explored laboratory-based albedo characterization, no agreement has been reached regarding the testing apparatus or conditions. In this work, a parametric study was conducted to develop a laboratory procedure. Special attention was given to the type of lighting and sensor sensitivity in order to accurately replicate solar illumination and ensure reliable measurements. Colorimetry was used to confirm material consistency across laboratory and field scales. The final transformation equation proposed is simple to apply and sufficiently accurate to provide reliable predictions of outdoor albedo
Pocket vane and penetrometer as quality control tool for Extrusion 3D concrete printing
No full textInternational audienceThe development of 3D concrete printing has led to a broad array of digital construction processes, with extrusion-based techniques emerging as the most widespread. Despite significant advances, the lack of standardized procedures for quality control and structural design continues to impede broader industrial adoption. Effective quality control in 3D concrete printing requires that the rheological properties of fresh materials align with the demands of the printing process, thereby calling for rapid, field-deployable testing methods. Among recent developments, the slug test has shown promise by enabling estimation of the initial shear yield stress based on the mass of extruded slugs. However, this method provides limited insight into buildability, which depends on the time-dependent evolution of the material’s yield stress. To address this gap, pocket vane and penetration tests have been adapted for use in 3D printing contexts. Vane tests, when conducted with pocket devices, allow in situ tracking of the yield stress over time, while penetration tests—performed using either slow or fast protocols—offer valuable information on the material’s strength development. This paper presents a protocol that combines pocket vane and penetrometer testing to simultaneously evaluate shear yield stress and internal friction angle in printable mortars. The proposed approach is validated through experimental investigations on four different mortar formulations, demonstrating its relevance for practical quality control in extrusion-based 3D concrete printing
Modélisation multiphysique du soudage par résistance par point
No full textInternational audienceModélisation multiphysique du soudage par résistance par poin
Développement d'un modèle numérique pour l'étude de structures composites hybrides à fibres de lin/carbone
No full textInternational audienceDéveloppement d'un modèle numérique pour l'étude de structures composites hybrides à fibres de lin/carbon
Les outils traditionnels au service de la protection des acteurs économiques face au numérique (volet 1)
No full textInternational audienc
From Data to Decision: A Digital Twin Supporting Sustainable Multi-Criteria Decision-Making for Heritage Territories and Their Users
No full textInternational audienceAs tourism pressure intensifies on UNESCO-listed heritage sites, innovative approaches are needed to reconcile visitor management with cultural preservation. This paper explores the potential of the territorial digital twin as a promising, yet still under development, instrument for sustainable governance. Within the ongoing JUNITER project, a dynamic digital twin is being developed for the Morbihan megalithic landscapes, recently inscribed on the UNESCO World Heritage list, progressively integrating heterogeneous data sources. This evolving model is designed to enable near-real-time simulation of visitor flows, early detection of saturation risks, and scenario-based support for strategic planning and heritage conservation. By combining behavioural modelling with multi-criteria decision analysis, JUNITER aims to provide a methodological foundation that could, in future, help assess and improve the sustainability, resilience, and robustness of heritage site management
Une mesure de suspension de fonctions peut être prise afin de préserver les conditions de travail au sein d'un service ainsi que les usagers de celui-ci
No full textInternational audienceNote sous CE, 23 janvier 2025, no 494959 (inédit au Lebon