HAL Portal UTC Université de Technologie de Compiègne
Not a member yet
11652 research outputs found
Sort by
Uncertainty quantification in regression neural networks using likelihood-based belief functions
No full textInternational audienceWe introduce a new method for quantifying prediction uncertainty in regression neural networks using evidential likelihood-based inference. The method is based on the Gaussian approximation of the likelihood function and the linearization of the network output with respect to the weights. Prediction uncertainty is described by a random fuzzy set inducing a predictive belief function. Preliminary experiments suggest that the approximations are very accurate and that the method allows for conservative uncertainty-aware predictions
Structural and biochemical characterization of SmoPG1, an exo-polygalacturonase from Selaginella moellendorffii
No full textInternational audiencePolygalacturonases (PGs) can modulate chemistry and mechanical properties of the plant cell wall through the degradation of pectins, one of its major constituents. PGs are largely used in food, beverage, textile, and paper industries to increase processes' performances. To improve the use of PGs, knowledge of their biochemical, structural and functional features is of prime importance. Our study aims at characterizing SmoPG1, a polygalacturonase from Selaginella moellendorffii, that belongs to the lycophytes. Transcription data showed that SmoPG1 was mainly expressed in S. moellendorffii shoots while phylogenetic analyses suggested that SmoPG1 is an exo-PG, which was confirmed by the biochemical characterization following its expression in heterologous system. Indeed, LC-MS/MS oligoprofiling using various pectic substrates identified galacturonic acid (GalA) as the main hydrolysis product. We found that SmoPG1 was most active on polygalacturonic acid (PGA) at pH 5, and that its activity could be modulated by different cations (Ca2+, Cu2+, Fe2+, Mg2+, Mn2+, Na2+, Zn2+). In addition, SmoPG1 was inhibited by green tea catechins, including (−)-epigallocatechin-3-gallate (EGCG). Docking analyses and MD simulations showed in detail amino acids responsible for the SmoPG1-EGCG interaction. Considering its expression yield and activity, SmoPG1 appears as a prime candidate for the industrial production of GalA
Optimal position and dimensions of embedded normal piezoelectric transducers, higher order plate models and experimental approach
No full textInternational audienceThis study investigates the different optimal parameters of piezoelectric transducers polarized in the thickness direction. The studied parameters are regarding sensor and actuator embedding position in the thickness, sensor ratio and sensor dimensions. It involves a parametric study using MUL2 FE numerical tool, via coupled electro-mechanical two-dimensional models. The performance of the corresponding configuration was judged by the maximum output electric potential signal induced in the sensor due to the propagated Lamb Wave induced by the actuator. An experimental campaign was carried out on glass fiber/ polyester plates embedded with piezoelectric transducers to validate the different results
Reliable Risk Assessment and Management using Probabilistic Fusion of Predictive Inter-Distance Profile for Urban Autonomous Driving
No full textInternational audienceAutonomous driving in urban scenarios has become more challenging due to the increase in Personal Light Electric Vehicles (PLEVs). PLEVs correspond mostly to electric devices such as gyropods and scooters. They exhibit varying velocity profiles as a result of their high acceleration capacity. Multiple hypotheses about their possible motion make autonomous driving very difficult, leading to the highly conservative behavior of most control algorithms. This paper proposes to solve this problem by performing a continuous risk assessment using a Fusion of Predictive Inter-Distance Profile (F-PIDP). Then a stochastic MPC algorithm performs effective risk management using the F-PIDP while taking into account adaptive constraints. The advantages of the proposed approach are demonstrated through simulations of multiple scenarios
A novel in-situ synthesized bioactive TaZrTi alloy by high energy laser beam for total knee arthroplasty application
No full textInternational audienceTotal knee arthroplasty is the gold standard treatment for end-stage knee osteoarthritis, which could significantly improve function and relieve pain for patients. However, the present materials and manufacturing approaches are dissatisfactory for knee implants. In the present work, a Ta60-Zr20-Ti20 (in wt%) coating was in-situ synthesized onto Ti6Al4V alloy via a high energy laser beam. The microstructure analysis indicated that α-Ta/ZrTi solid solution generated into body-centered cubic (bcc) grains in dimensions of several to hundred micrometers. The Ta-riched dendrites nucleated, grew, and expanded in the ZrTi matrix. The mechanical and biological properties of the in-situ synthesized TaZrTi sample are compared with the conventionally rolled Ti6Al4V, which is also the substrate alloy. The mean values of hardness and elastic modulus of in-situ synthesized TaZrTi tested by the nanoindenter, are 10.31 and 134.84 GPa, respectively. These results are harder than the Ti6Al4V counterpart (5.14 GPa) with a preferred slightly low elastic modulus (152.95 GPa of Ti6Al4V). Subsequently, the in-situ synthesized TaZrTi exhibits an excellent wear resistance demonstrated by nano-scratch experiments. In addition, the osteoblastic and endothelial cells were cultured in vitro to evaluate the biocompatibility of in-situ synthesized TaZrTi. The results show a comparable bioactivity and surface affinity compared with the rolled Ti6Al4V in 7 days. Briefly, this in-situ synthesized TaZrTi alloy is expected to shed light on load-bearing orthopedic applications to relieve the pain of patients, and the findings in this study provide an insightful understanding of laser additive manufactured TaZrTi alloy. © 2023 Elsevier B.V
Impact resisting mechanism of tension-torsion coupling metamaterials
No full textInternational audienceTension–torsion coupling (TTC) metamaterials are man-made architectures that exhibit unexpected rotational deformation when subjected to unidirectional loads. Since their discovery, numerous studies have been conducted to investigate the static properties of TTC metamaterials resulting from this twisting effect. However, research on their dynamic properties, particularly in terms of impact resistance, has been limited with little understanding of the underlying mechanisms. In this study, we performed simulations at different length scales to quantitatively analyze the energy flow during impact and gained insights into the mechanisms of energy damping and impact resistance. Our simulations demonstrate that the twisting effect of cellular materials generally leads to reduced stiffness, which is advantageous for improving impact mitigation. Furthermore, we discovered that the unique chiral characteristics of TTC metamaterials enable a greater amount of strain energy to be stored during impact, and this portion of energy will be ultimately dissipated through after-impact vibrations. Additionally, we found that energy dissipation via friction between lateral struts or ligaments, although less significant, is enhanced by the TTC effect. The current study provides a preliminary analysis of energy flow and sheds light on the mechanisms underlying the energy damping and impact resistance of TTC metamaterials. These findings not only contribute to a better understanding of the dynamic properties of TTC metamaterials but also have practical implications for the advanced design of impact resistance structures
Algebraic Product Is the Only "And-like" Operation for Which Normalized Intersection Is Associative: A Proof
No full textMachine Learning and Other Soft Computing Techniques: Biomedical and Related Applications - Studies in Systems, Decision and ControlInternational audienceFor normalized fuzzy sets, intersection is, in general, not normalized. So, if we want to limit ourselves to normalized fuzzy sets, we need to normalize the intersection. It is known that for algebraic product, the normalized intersection is associative, and that for many other ``and''-operations (t-norms), normalized intersection is not associative. In this paper, we prove that algebraic product is the only ``and''-operation for which normalized intersection is associative
Recognizing Similar Crises through the Application of Ontology-based Knowledge Mining
No full textInternational audienceRecognizing and learning from similar crisis situations is crucial for the development of effective response strategies. This study addresses the challenge of identifying similarities within a wide range of crisis-related information. To overcome this challenge, we employed an ontology-based crisis situation knowledge base enriched with crisis-related information. Additionally, we implemented a semantic similarity measure to assess the degree of similarity between crisis situations. Our investigation specifically focuses on recognizing similar crises through the application of ontology-based knowledge mining. Through our experiments, we demonstrate the accuracy and efficiency of our approach to recognizing similar crises. These findings highlight the potential of ontology-based knowledge mining for enhancing crisis recognition processes and improving overall crisis management strategies
Extremely tortuous sound absorbers with labyrinthine channels in non-porous and microporous solid skeletons
No full textInternational audienceAn assembly of additively-manufactured modules to form two-dimensional networks of labyrinthine slits results in a sound absorber with extremely high tortuosity and thereby a relatively low-frequency quarter wavelength resonance. Fully analytical modelling is developed for the generic design of such composite acoustic panels, allowing rapid exploration of various specific designs. In addition to single-porosity panels, a double porosity case is also investigated, in which the skeleton of the composite is microporous with a permeability very much lower than that of the tortuous slit network. The analytical modelling is verified by various numerical calculations, as well as sound absorption measurements performed on several 3D printed samples of modular composite panels. The experimental validation required overcoming the nontrivial difficulties related to additive manufacturing and testing samples of extreme tortuosity. On the other hand, due to the substantial two-dimensionality and modularity of the proposed design, such absorbers can be manufactured by traditional methods suitable for mass production of large-surface panels. The experimental results fully confirmed the theoretical predictions that significant sound absorption, almost perfect at the peak, can be achieved at relatively low frequencies using very thin panels, especially those with double porosity
Experimental observation of exceptional points in coupled pendulums
No full textInternational audienc