Search Results (979)

Battery electric vehicles (BEVs) are gaining market share, yet range anxiety and sparse charging still create demand for hybrids with combustion-engine range extenders. Range-extender vehicles face high customer expectations for noise, vibration, and harshness (NVH) due to their direct comparability with fully electric vehicles. Key challenges include the vibrations of the internal combustion engine, especially from vehicle-induced starts, and the discontinuous operating principle. A technological concept to reduce vibrations in the drivetrain and on the engine mounts, called “FEVcom,” relies on rolling moment compensation. In this concept, a counter-rotating electric machine is coupled to the internal combustion engine via a gear stage to minimize external mount forces. However, due to high speed fluctuations of the crankshaft, the gear drive tends to rattle, which is perceived as disturbing and must be avoided. As part of this work, the rolling moment compensation system was examined regarding its vibration excitation, and an extension to prevent gear rattling was simulated and optimized. For the simulation, the extension, based on a chain or belt drive, was set up as a multi-body simulation model in combination with the range extender and examined dynamically at different speeds. Variations of the extended system were simulated, and recommendations for an optimized layout were derived. This work demonstrates the feasibility of successful rattling avoidance in a range-extender drivetrain with full utilization of the rolling moment compensation. It also provides a solid foundation for further detailed investigations and for developing a prototype for experimental validation based on the understanding gained of the system. Full article

For electric vehicles widely used recently, lithium-ion batteries serve as the primary energy storage units, affecting the vehicles’ performance, safety, and lifespan. Accurate state of charge (SOC) estimation is pivotal for the battery management system (BMS) to enhance the predictability of the vehicle’s range and avert thermal runaway due to improper charging methods. In this study, an adaptive SOC estimation methodology was developed using parameter identification with forgetting factor recursive least squares (FFRLS). These parameters are then incorporated into a dual adaptive extended Kalman filter (DAEKF) for SOC estimation under varying load conditions. DAEKF is used to dynamically adjust the covariance matrices for process and measurement noises, significantly enhancing the filter’s adaptability and precision. The integration of FFRLS and DAEKF enables a robust SOC estimation of electric vehicles, featuring rapid computation speeds, high accuracy, and excellent adaptability, positioning them as ideal candidates for enhancements in battery management system technology. Full article

Miller’s rule originated as an empirical relation between the nonlinear and linear optical coefficients of materials. It is now accepted as a useful tool for guiding experiments and computational materials discovery, but its theoretical foundation had long been limited to a derivation for the classical Lorentz model with a weak anharmonic perturbation. Recently, we developed a mathematical framework which enabled us to prove that Miller’s rule is equally valid for quantum anharmonic oscillators, despite different dynamics due to zero-point fluctuations and further quantum-mechanical effects. However, our previous derivation applied only to one-dimensional oscillators and to the special case of second- and third-harmonic generation in a monochromatic electric field. Here we extend the proof to three-dimensional quantum anharmonic oscillators and also treat all orders of the nonlinear response to an arbitrary multi-frequency field. This makes the results applicable to a much larger range of physical systems and nonlinear optical processes. The obtained generalized Miller formulae rigorously express all tensor elements of the frequency-dependent nonlinear susceptibilities in terms of the linear susceptibility and thus allow a computationally inexpensive quantitative prediction of arbitrary parametric frequency-mixing processes from a small initial dataset. Full article

Recycled asphalt pavement (RAP) can support traffic loads comparable to those of roads constructed with conventional materials. The structural evaluation of RAP is performed through the deflection generated by vehicles via recoverable deflection in the pavement layers. The deflection record is translated into a curve that geometrically interprets the behavior of the layers that make up the pavement. In this study, a falling weight deflectometer (FWD) was used to emulate transit loads and measure deflection in two models. Both contained 30% RAP, and one of them had fiberglass geogrid in the center of the asphalt layer. Through normalized maximum deflection (limit value based on constant stress), the structural index (SI), and the dynamic stiffness modulus (DSM), the structural behavior of the models under different load levels was evaluated. The pavement structure exhibited similarities in strength for both models subjected to impact. The presence of the geogrid reinforcement (Z1) showed structural index values ranging between 0.17 and 0.54, while the layer without geogrid (Z2) presented structural index values in a range of 0.23 to 0.78. In addition, the dynamic stiffness modulus presented a difference of 10 kN/mm between the maximums of the models in favor of reinforcement with glass fiber geogrid. Therefore, low structural index values are associated with the interaction between RAP and geogrid, highlighting this combination as an innovative and functional system for road surfaces, while the dynamic stiffness modulus indicates the stability and structural integrity of sustainable pavement, which has the potential to extend its lifespan. Full article

This paper investigates the time-varying efficiency of Western European stock markets and examines how macroeconomic events defined as endogenous and exogenous shocks influence the degree of efficiency by either long-range dependence or mean reverting. We apply a rolling-window detrended fluctuation analysis (DFA) with two window sizes, complemented by the Efficiency Index to synthetize multiple measures of market efficiency. The results confirm that efficiency evolves dynamically in response to macroeconomic disruptions. Specifically, endogenous shocks tend to generate anti-persistent behavior, while exogenous shocks are associated with long-memory effect. These shifts in efficiency are also reflected in rolling Kurtosis estimates, suggesting that only the most severe shocks produce spikes in Kurtosis, fat-tailed returns distributions, and structural inefficiencies. This dual approach allows us to classify shocks as major or minor based on their joint impact on both market efficiency and tail behavior. Overall, our findings support the adaptive market hypothesis and extend its implications through the fractal market hypothesis by underlining the role of heterogenous investment horizons during periods of turmoil. The combined use of dynamic DFA and Kurtosis offer a framework to assess how financial markets adapt to different types of macroeconomic shocks. Full article

The pursuit of universal, symmetric semantic representations within large language models (LLMs) faces a fundamental challenge: the inherent asymmetry of natural languages. Different languages exhibit vast disparities in syntactic structures, lexical choices, and cultural nuances, making the creation of a truly shared, symmetric embedding space a non-trivial task. This paper aims to address this critical problem by introducing a novel framework to forge robust and symmetric multilingual sentence embeddings. Our approach, named DACL (Dynamic Asymmetric Contrastive Learning), is anchored in two powerful asymmetric learning paradigms: Contrastive Learning and Dynamic Curriculum Learning (DCL). We extend Contrastive Learning to the multilingual context, where it asymmetrically treats semantically equivalent sentences from different languages (positive pairs) and sentences with distinct meanings (negative pairs) to enforce semantic symmetry in the target embedding space. To further refine this process, we incorporate Dynamic Curriculum Learning, which introduces a second layer of asymmetry by dynamically scheduling training instances from easy to hard. This dual-asymmetric strategy enables the model to progressively master complex cross-lingual relationships, starting with more obvious semantic equivalences and advancing to subtler ones. Our comprehensive experiments on benchmark cross-lingual tasks, including sentence retrieval and cross-lingual classification (XNLI, PAWS-X, MLDoc, MARC), demonstrate that DACL significantly outperforms a wide range of established baselines. The results validate our dual-asymmetric framework as a highly effective approach for forging robust multilingual embeddings, particularly excelling in tasks involving complex linguistic asymmetries. Ultimately, this work contributes a novel dual-asymmetric learning framework that effectively leverages linguistic asymmetry to achieve robust semantic symmetry across languages. It offers valuable insights for developing more capable, fair, and interpretable multilingual LLMs, emphasizing that deliberately leveraging asymmetry in the learning process is a highly effective strategy. Full article

Plant disease segmentation has achieved significant progress with the help of artificial intelligence. However, deploying high-accuracy segmentation models in resource-limited settings faces three key challenges, as follows: (A) Traditional dense attention mechanisms incur quadratic computational complexity growth (O($n^{2}d$)), rendering them ill-suited for low-power hardware. (B) Naturally sparse spatial distributions and large-scale variations in the lesions on leaves necessitate models that concurrently capture long-range dependencies and local details. (C) Complex backgrounds and variable lighting in field images often induce segmentation errors. To address these challenges, we propose Sparse-MoE-SAM, an efficient framework based on an enhanced Segment Anything Model (SAM). This deep learning framework integrates sparse attention mechanisms with a two-stage mixture of experts (MoE) decoder. The sparse attention dynamically activates key channels aligned with lesion sparsity patterns, reducing self-attention complexity while preserving long-range context. Stage 1 of the MoE decoder performs coarse-grained boundary localization; Stage 2 achieves fine-grained segmentation by leveraging specialized experts within the MoE, significantly enhancing edge discrimination accuracy. The expert repository—comprising standard convolutions, dilated convolutions, and depthwise separable convolutions—dynamically routes features through optimized processing paths based on input texture and lesion morphology. This enables robust segmentation across diverse leaf textures and plant developmental stages. Further, we design a sparse attention-enhanced Atrous Spatial Pyramid Pooling (ASPP) module to capture multi-scale contexts for both extensive lesions and small spots. Evaluations on three heterogeneous datasets (PlantVillage Extended, CVPPP, and our self-collected field images) show that Sparse-MoE-SAM achieves a mean Intersection-over-Union (mIoU) of 94.2%—surpassing standard SAM by 2.5 percentage points—while reducing computational costs by 23.7% compared to the original SAM baseline. The model also demonstrates balanced performance across disease classes and enhanced hardware compatibility. Our work validates that integrating sparse attention with MoE mechanisms sustains accuracy while drastically lowering computational demands, enabling the scalable deployment of plant disease segmentation models on mobile and edge devices. Full article

Toward the innovative design of tunable structures for energy generation, this paper presents an extended Floating Frame of Reference (FFR) formulation capable of modeling slope discontinuities in flexible multibody systems—overcoming a key limitation of conventional FFR methods that assume slope continuity. The model is validated using a spatial double-pendulum structure composed of circular beam elements, representative of out-of-plane energy harvesting systems. To investigate the influence of boundary constraints on dynamic behavior, three electromagnetic clamping configurations—Fixed–Free–Free (XFF), Fixed–Free–Fixed (XFX), and Free–Fixed–Free (FXF)—are implemented. Tri-axial accelerometer measurements are analyzed via Fast Fourier Transform (FFT), revealing natural frequencies spanning from 38.87 Hz (lower frequency range) to 149.01 Hz (higher frequency range). For the lower frequency range, the FFR results (38.76 Hz) show a close match with the experimental prediction (38.87 Hz) and ANSYS simulation (36.49 Hz), yielding 0.28% error between FFR and experiments and 5.85% between FFR and ANSYS. For the higher frequency range, the FFR model (148.17 Hz) achieves 0.56% error with experiments (149.01 Hz) and 0.85% with ANSYS (146.91 Hz). These high correlation percentages validate the robustness and accuracy of the proposed FFR formulation. The study further shows that altering boundary conditions enables effective frequency tuning in discontinuous structures—an essential feature for the optimization of application-specific systems such as wave energy converters. This validated framework offers a versatile and reliable tool for the design of vibration-sensitive devices with geometric discontinuities. Full article

Anthropogenic climate change threatens production of essential natural resources, including food, fiber, and fresh water, and provisioning of ecosystem services such as carbon sequestration, increasing the risk of societal collapse. The Human and Nature Dynamics (HANDY) model simulates the effect of resource overexploitation on societal collapse but lacks representation of feedbacks between climate change and resource regeneration in ecological systems. We extend the HANDY model by integrating models of climate change and ecological function to examine the risk of societal collapse. We conducted a sensitivity analysis of our expanded model by systematically varying key parameters to examine the range of plausible socio-ecological conditions and evaluate model uncertainty. We find that lowered greenhouse gas emissions and resilient ecosystems can delay societal collapse by up to approximately 500 years, but that any scenario with greater than net-zero greenhouse gas emissions ultimately leads to societal collapse driven by climate-induced loss of ecosystem function. Reductions in greenhouse gas emissions are the most effective intervention to delay or prevent societal collapse, followed by the conservation and management of resilient ecological systems to sequester atmospheric carbon. Full article

Human keypoint detection has become a fundamental task in computer vision, underpinning a wide range of downstream applications such as action recognition, intelligent surveillance, and human–computer interaction. Accurate localization of keypoints is crucial for understanding human posture, behavior, and interactions in various environments. In this paper, we propose a deep-learning-based human skeletal keypoint detection framework that leverages a High-Resolution Network (HRNet) to achieve robust and precise keypoint localization. Our method maintains high-resolution representations throughout the entire network, enabling effective multi-scale feature fusion, without sacrificing spatial details. This approach preserves the fine-grained spatial information that is often lost in conventional downsampling-based methods. To evaluate its performance, we conducted extensive experiments on the COCO dataset, where our approach achieved competitive performance in terms of Average Precision (AP) and Average Recall (AR), outperforming several state-of-the-art methods. Furthermore, we extended our pipeline to support multi-person keypoint detection in real-time scenarios, ensuring scalability for complex environments. Experimental results demonstrated the effectiveness of our method in both single-person and multi-person settings, providing a comprehensive and flexible solution for various pose estimation tasks in dynamic real-world applications. Full article

This study investigates the concentrations and patterns of dioxins and dioxin-like PCBs (TEQ Diox+PCB-DL) and non-dioxin-like PCBs (PCB-NDL) in eels from Lake Garda, assessing their relationship with biometric and lipid parameters. TEQ Diox+PCB-DL levels ranged from 1.70 to 77.1 pg/g (median: 9.90 pg/g), while PCB-NDL levels spanned from 14.0 to 1620 ng/g (median: 65.5 ng/g). Significant, albeit low, correlations were found: length and weight were negatively correlated, and lipid content was positively correlated, with both contaminants. Multivariable regression confirmed length and lipid percentage as significant predictors, although the models explained a limited proportion of variance (R²: 0.23 and 0.17). Classification-based analyses showed that irregularly contaminated eels were shorter and had a higher lipid content. Multinomial logistic regression supported these findings, but showed limited predictive accuracy (AUC = 0.63). Notably, 28 out of 90 samples exceeded the EU regulatory limit for TEQ Diox+PCB-DL, and several surpassed the threshold for PCB-NDL, highlighting potential public health risks. Given the lipophilic nature and toxicity of these compounds, continued monitoring is warranted. The findings underscore the need for broader environmental assessments to better understand pollutant dynamics and support regulatory actions, including the extended ban on eel fishing in the region. Full article

This study introduces a memristive Hopfield neural network (M-HNN) model to investigate electromagnetic radiation impacts on neural dynamics in complex electromagnetic environments. The proposed framework integrates a magnetic flux-controlled memristor into a three-neuron Hopfield architecture, revealing significant alterations in network dynamics through comprehensive nonlinear analysis. Numerical investigations demonstrate that memristor-induced electromagnetic effects induce distinctive phenomena, including coexisting attractors, transient chaotic states, symmetric bifurcation diagrams and attractor structures, and constant chaos. The proposed system can generate more than 12 different attractors and extends the chaotic region. Compared with the chaotic range of the baseline Hopfield neural network (HNN), the expansion amplitude reaches 933%. Dynamic characteristics are systematically examined using phase trajectory analysis, bifurcation mapping, and Lyapunov exponent quantification. Experimental validation via a DSP-based hardware implementation confirms the model’s operational feasibility and consistency with numerical predictions, establishing a reliable platform for electromagnetic–neural interaction studies. Full article

This study presents an advanced simulation-based methodology for evaluating the dynamic vibrational behavior of the operator–vehicle assembly in autonomous electric agricultural tractors. Using the TE-0 electric tractor as the experimental platform, the research is structured into three integrated stages. In the first stage, a seated anthropometric virtual model of the human operator is developed based on experimental data and biomechanical validation. The second stage involves a detailed modal analysis of the TE-0 electric tractor using Altair Sim Solid, with the objective of determining the natural frequencies and vibration modes in the [0–80] Hz range, in compliance with ISO 2631-1. This analysis captures both the structural-induced frequencies—associated with the chassis, wheelbase, and metallic frame—and the operational-induced frequencies, influenced by the velocity and terrain profile. Subsequently, the modal analysis of the “Grammer Cabin Seat” is conducted to assess its dynamic response and identify critical vibration modes, highlighting how the seat behaves under vibrational stimuli from the tractor and terrain. The third stage extends the analysis to the virtual operator model seated on the tractor seat, investigating the biomechanical response of the human body and the operator–seat–vehicle interaction during simulated motion. Simulations were carried out using SolidWorks 2023 and Altair Sim Solid over a frequency range of [0–80] Hz, corresponding to operation on unprocessed soil covered with grass, at a constant forward speed of 7 km/h. The results reveal critical resonance modes and vibration transmission paths that may impact operator health, comfort, and system performance. The research contributes to the development of safer, more ergonomic, and sustainable autonomous agricultural transport systems. By simulating real-world operation scenarios and integrating a rigorously validated experimental protocol—including vibration data acquisition, biomechanical modeling, and multi-stage modal analysis—this study demonstrates the importance of advanced modeling in optimizing system-level performance, minimizing harmful vibrations, and supporting the transition toward resilient and eco-efficient electric tractor platforms in smart agricultural mobility. Full article

High dynamic range (HDR) tone mapping techniques have been widely studied to effectively represent the broad dynamic range of real-world scenes. However, generating an HDR image from multiple low dynamic range (LDR) images captured at different exposure levels can introduce ghosting artifacts in dynamic scenes. Moreover, methods that estimate HDR information from a single LDR image often suffer from inherent accuracy limitations. To overcome these limitations, this study proposes a novel image processing technique that extends the dynamic range of a single LDR image. This technique achieves the goal through leveraging a Convolutional Neural Network (CNN) to generate a synthetic Near-Infrared (NIR) image—one that emulates the characteristic of real NIR imagery being less susceptible to diffraction, thus preserving sharper outlines and clearer details. This synthetic NIR image is then fused with the original LDR image, which contains color information, to create a tone-distributed HDR-like image. The synthetic NIR image is produced using a lightweight U-Net-based autoencoder, where the encoder extracts features from the LDR image, and the decoder synthesizes a synthetic NIR image that replicates the characteristics of a real NIR image. To enhance feature fusion, a cardinality structure inspired by Extended-Efficient Layer Aggregation Networks (E-ELAN) in You Only Look Once Version 7 (YOLOv7) and a modified convolutional block attention module (CBAM) incorporating a difference map are applied. The loss function integrates a discriminator to enforce adversarial loss, while VGG, structural similarity index, and mean squared error losses contribute to overall image fidelity. Additionally, non-reference image quality assessment losses based on BRISQUE and NIQE are incorporated to further refine image quality. Experimental results demonstrate that the proposed method outperforms conventional HDR techniques in both qualitative and quantitative evaluations. Full article

This study investigated the thermomechanical behaviour of the nozzle of a GTM 400MOD miniature turbojet engine during combustion of aviation kerosene and co-combustion of kerosene with hydrogen. Numerical analysis was based on experiments conducted on a dedicated test rig at engine speeds ranging from 31,630 rpm to 65,830 rpm, providing data on the temperature and dynamic pressure at the nozzle outlet. These data served as input to numerical analyses using the ANSYS Fluent, Steady-State Thermal, and Static Structural modules to evaluate exhaust gas flow, temperature distribution, and stress and strain states. The paper performed a basic analysis with additional simplifications, and an extended analysis that took into account, among other things, thermal radiation in the flow. The results of the basic analysis show that, at comparable thrust levels, co-firing and pure kerosene combustion yield similar nozzle temperature distributions, with maximum wall temperatures ranging from 978 K to 1090 K, which remain below the allowable limit of 1193 K (920 °C). Maximum stresses reached approximately 261 MPa, close to but not exceeding the yield strength of 316 stainless steel. Maximum nozzle deformation did not exceed 0.8 mm. Small dynamic pressure fluctuations were observed; For example, at 31,630 rpm, co-firing increased the maximum dynamic pressure from 1.56 × 104 Pa to 1.63 × 104 Pa, while at 47,110 rpm, it decreased from 4.05 × 104 Pa to 3.89 × 104 Pa. The extended analysis yielded similar values for the nozzle temperature and pressure distributions. Stress and strain increased by more than 76% and 78%, respectively, compared to the baseline analysis. The results confirm that hydrogen co-firing does not significantly alter the nozzle thermomechanical loads, suggesting that this emission-free fuel can be used without negatively impacting the nozzle’s structural integrity under the tested conditions. The methodology, combining targeted experimental measurements with coupled CFD and FEM simulations, provides a reliable framework for assessing material safety margins in alternative fuel applications in small turbojet engines. Full article