Search Results (123)

This study presents a multiscale, uncertainty-aware hybrid deep learning approach addressing the short-term wind speed prediction problem, which is critical for the reliable planning and operation of wind energy systems. Wind signals are decomposed using adaptive variational mode decomposition (VMD), and the resulting wind components are processed together with meteorological data through a dual-stream CNN–BiLSTM architecture. Based on this multiscale representation, probabilistic forecasts are generated using quantile regression to capture best- and worst-case scenarios for decision-making purposes. Unlike fixed prediction intervals, the proposed approach produces adaptive prediction bands that expand during unstable wind conditions and contract during calm periods. The developed model is evaluated using four years of meteorological data from the Afyonkarahisar region of Türkiye. While the proposed model achieves competitive point forecasting performance (RMSE = 0.700 m/s and MAE = 0.54 m/s), its main contribution lies in providing reliable probabilistic forecasts through well-calibrated uncertainty quantification, offering decision-relevant information beyond single-point predictions. The proposed method is compared with a classical CNN–LSTM and several structural variants. Furthermore, SHAP-based explainability analysis indicates that seasonal and solar-related variables play a dominant role in the forecasting process. Full article

This paper introduces a fractional generalization of the classical Legendre differential equation based on the Atangana–Baleanu–Caputo (ABC) derivative. A novel fractional Legendre-type operator is rigorously defined within a functional framework of continuously differentiable functions with absolutely continuous derivatives. The associated initial value problem is reformulated as an equivalent Volterra integral equation, and existence and uniqueness of classical solutions are established via the Banach fixed-point theorem, supported by a proved Lipschitz estimate for the ABC derivative. A constructive solution representation is obtained through a Volterra–Neumann series, explicitly revealing the role of Mittag–Leffler functions. We prove that the fractional solutions converge uniformly to the classical Legendre polynomials as the fractional order approaches unity, with a quantitative convergence rate of order $O(1-\alpha )$ under mild regularity assumptions on the Volterra kernel. A fully reproducible quadrature-based numerical scheme is developed, with explicit kernel formulas and implementation algorithms provided in appendices. Numerical experiments for the quadratic Legendre mode confirm the theoretical convergence and illustrate the smooth interpolation between fractional and classical regimes. An application to time-fractional diffusion in spherical coordinates demonstrates that the operator arises naturally in physical models, providing a mathematically consistent tool for extending classical angular analysis to fractional settings with memory. Full article

Pixelwise dense representations are more prevalent in the field of iris recognition, also known as iris templates or IrisCodes. Almost all previous works of this kind are deterministic. To be specific, pixel-level representations are exclusively derived from certain point-by-point modeling, including filter responses, phase correlations, and ordinal relations. Moreover, the binary mask indicating valid iris regions is solely determined by a fixed threshold or the output of standalone segmentation and localization algorithms. Uncertainty in acquisition factors in the process of iris imagery formation is not considered. In this paper, we propose a simple yet effective plug-and-play building block termed dual dense uncertainty embedding (D²UE), which can be seamlessly incorporated into deep learning (DL) frameworks that extract dense representations for iris recognition. D²UE has two pathways wherein both take dense feature maps of the backbone network as input. One pathway of D²UE predicts a variance-scaling map (VSM) and then applies it to an adaptive threshold-masking operation on the iris image. The dynamic threshold for each pixel in this manner is dependent on not only the intensity distribution of the iris image but also each pixel’s low-level uncertainty. The other pathway of D²UE adopts an over-parameterization technique and extracts uncertainty-embedded dense representations (UEDRs) by modeling each pixel’s contextual uncertainty. Extensive experiments on several iris datasets demonstrate that recognition performance under both within-database and cross-database settings can be significantly improved by incorporating D²UE into the baseline method. By integrating D²UE into various deep learning frameworks and evaluating their performance across multiple datasets, the results demonstrate that D²UE can be seamlessly incorporated into diverse architectures and can significantly enhance their recognition capabilities. D²UE only incurs slight computational overhead while surpassing a few SOTA methods with a large backbone network and much more training budget. Full article

This study investigates lump wave structures that arise from the interplay of dispersion and nonlinearity in a generalized Calogero–Bogoyavlenskii–Schiff-like model with spatially symmetric nonlinearity in (2+1) dimensions. A generalized bilinear representation of the governing equation is formulated using extended bilinear derivatives of the fourth order, providing a convenient framework for analytic treatment. Through symbolic computation, we construct positive quadratic wave solutions, which give rise to rationally localized lump wave tructures that decay algebraically in all spatial directions at fixed time. Analysis shows that the critical points of these quadratic waves lie along a straight line in the spatial plane and propagate at a constant velocity. Along this characteristic trajectory, the amplitudes of the lump waves remain essentially unchanged, reflecting the stability of these coherent structures. The emergence of these lumps is primarily driven by the combined influence of five dispersive terms in the model, highlighting the crucial role of higher-order dispersion in balancing the nonlinear interactions and shaping the resulting localized waveforms. Full article

This paper presents an innovative approach to the modeling and dynamic analysis of DC–DC converters in photovoltaic applications. Departing from traditional studies that focus on the transfer function from duty cycle to output voltage, this work investigates the duty cycle to input voltage transfer function, which is critical for accurate dynamic representation of photovoltaic systems. A notable contribution of this study is the integration of the PV panel behavior in the small-signal representation, considering a model-derived differential resistance for various operating points. This technique enhances the model’s accuracy across different operating regions. The paper also validates the effectiveness of this linearization method through small-signal analysis. A comprehensive comparison is conducted among several non-isolated converter topologies such as Boost, Buck–Boost, Ćuk, and SEPIC under both open-loop and closed-loop conditions. To ensure fairness, all converters are designed using a consistent set of constraints, and controllers are tuned to maintain similar phase margins and crossover frequencies across topologies. In addition, a gain-scheduling control strategy is implemented for the Boost converter, where the PI gains are dynamically adapted as a function of the PV operating point. This approach demonstrates superior closed-loop performance compared to a fixed controller tuned only at the maximum power point, further highlighting the benefits of the proposed modeling and control framework. This systematic study therefore provides an objective evaluation of dynamic performance and offers valuable insights into optimal converter architectures and advanced control strategies for photovoltaic systems. Full article

Neuromorphic hardware systems allow efficient implementation of artificial neural networks (ANNs) across various applications that demand high data throughput, reduced physical size, and low energy consumption. Field-Programmable Gate Arrays (FPGAs) possess inherent features that can be aligned with these requirements. However, implementing ANNs on FPGAs also presents challenges, including the computation of the neuron activation functions, due to the balance between resource constraints and numerical precision. This paper proposes a resource-efficient hardware approximation method for the sigmoid function, utilizing a combination of first- and second-degree polynomial functions. The method aims mainly to minimize the approximation error. This paper also evaluates the obtained results against existing techniques and discusses their significance. The experimental results showed that, although the proposed method mainly aimed to minimize the approximation error, it also had lower hardware resource usage than several of the most closely related works. Using 16-bit fixed-point number representation, the absolute mean error was $1.66\times {10}^{-3}$ by using 0.04% of the logic blocks and 3.21% of the DSP blocks in a Ciclone V 5CGXFC7C7F23C8 FPGA Device. Full article

With rapid urbanization and growing population inflows into metropolitan areas, crowd gatherings have become increasingly frequent and dense, posing significant challenges to public safety management. Although existing crowd gathering detection methods have achieved notable progress, they still face major limitations: most rely heavily on local texture or density features and lack the capacity to model contextual information, making them ineffective under severe occlusions and complex backgrounds. Additionally, fixed-scale feature extraction strategies struggle to adapt to crowd regions with varying densities and scales, and insufficient attention to densely populated areas hinders the capture of critical local features. To overcome these challenges, we propose a point-supervised framework named Multi-Scale Convolutional Attention Network (MSCANet). MSCANet adopts a context-aware architecture and integrates multi-scale feature extraction modules and convolutional attention mechanisms, enabling it to dynamically adapt to varying crowd densities while focusing on key regions. This enhances feature representation in complex scenes and improves detection performance. Extensive experiments on public datasets demonstrate that MSCANet achieves high counting accuracy and robustness, particularly in dense and occluded environments, showing strong potential for real-world deployment. Full article

The primary objective of this study is to provide a more precise and beneficial mathematical model for assessing corruption dynamics by utilizing non-local derivatives. This research aims to provide solutions that accurately capture the complexities and practical behaviors of corruption. To illustrate how corruption levels within a community change over time, a non-linear deterministic mathematical model has been developed. The authors present a non-integer order model that divides the population into five subgroups: susceptible, exposed, corrupted, recovered, and honest individuals. To study these corruption dynamics, we employ a new method for solving a time-fractional corruption model, which we term the q-homotopy analysis transform approach. This approach produces an effective approximation solution for the investigated equations, and data is shown as 3D plots and graphs, which give a clear physical representation. The stability and existence of the equilibrium points in the considered model are mathematically proven, and we examine the stability of the model and the equilibrium points, clarifying the conditions required for a stable solution. The resulting solutions, given in series form, show rapid convergence and accurately describe the model’s behaviour with minimal error. Furthermore, the solution’s uniqueness and convergence have been demonstrated using fixed-point theory. The proposed technique is better than a numerical approach, as it does not require much computational work, with minimal time consumed, and it removes the requirement for linearization, perturbations, and discretization. In comparison to previous approaches, the proposed technique is a competent tool for examining an analytical outcomes from the projected model, and the methodology used herein for the considered model is proved to be both efficient and reliable, indicating substantial progress in the field. Full article

This study introduces a new class of coupled differential systems described by fractal–fractional Caputo derivatives with both constant and state-dependent delays. In contrast to traditional delay differential equations, the proposed framework integrates memory effects and geometric complexity while capturing adaptive feedback delays that vary with the system’s state. Such a formulation provides a closer representation of biological and physical processes in which delays are not fixed but evolve dynamically. Sufficient conditions for the existence and uniqueness of solutions are established using fixed-point theory, while the stability of the solution is investigated via the Hyers–Ulam (HU) stability approach. To demonstrate applicability, the approach is applied to two illustrative examples, including a predator–prey interaction model. The findings advance the theory of fractional-order systems with mixed delays and offer a rigorous foundation for developing realistic, application-driven dynamical models. Full article

Quantization-aware training (QAT) has emerged as a key strategy for enabling efficient deep learning inference on resource-constrained platforms. Yet, most existing approaches rely on static, manually selected numeric formats—fixed-point or floating-point—and fixed bit-widths, limiting their adaptability and often requiring extensive design effort or architecture search. In this work, we introduce a novel QAT framework that breaks this rigidity by jointly learning, during training, both the numeric representation format and the associated bit-widths in an end-to-end differentiable manner. At the core of our method lies a unified parameterization that is capable of emulating both fixed- and floating-point arithmetic, paired with a bit-aware loss function that penalizes excessive precision in a hardware-aligned fashion. We demonstrate that our approach achieves state-of-the-art trade-offs between accuracy and compression on MNIST, CIFAR-10, and CIFAR-100, reducing average bit-widths to as low as 1.4 with minimal accuracy loss. Furthermore, FPGA implementation using Xilinx FINN confirms over $5\times$ LUT and $4\times$ BRAM savings. This is the first QAT method to unify numeric format learning with differentiable precision control, enabling highly deployable, precision-adaptive deep neural networks. Full article

Board gender diversity (BGD) has gained attention as a governance mechanism that may influence corporate financial decisions. However, empirical evidence from Latin America and the Caribbean (LAC) remains limited despite the region’s significant gender disparities in corporate leadership and distinct institutional characteristics. This study examines how BGD affects capital structure decisions in LAC firms, drawing on agency theory and resource dependency theory. We analyze a panel dataset of 403 firms from 2015 to 2022, sourced from the London Stock Exchange Group database, using fixed effects models with Driscoll–Kraay standard errors to control for firm heterogeneity and econometric concerns. Results show that BGD is significantly and negatively associated with leverage ratios, with a one percentage point increase in female board representation corresponding to a 0.15 to 0.25 percentage point decrease in debt-to-capital ratios. This relationship is robust across multiple specifications and exhibits threshold effects, with stronger impacts when female representation reaches 20% or higher. The negative association is more pronounced for larger firms, consistent with enhanced governance benefits in complex organizations. Our findings suggest that gender-diverse boards exercise more effective oversight of financial decisions, leading to more conservative capital structures in emerging markets where governance mechanisms are particularly important for firm credibility and stakeholder confidence. Full article

3D object detection is a crucial technology in fields such as autonomous driving and robotics. As a direct representation of the 3D world, point cloud data plays a vital role in feature extraction and geometric representation. However, in real-world applications, point cloud data often suffers from occlusion, resulting in incomplete observations and degraded detection performance. Existing methods, such as PG-RCNN, generate semantic surface points within each Region of Interest (RoI) using a single grid size. However, a fixed grid scale cannot adequately capture multi-scale features. A grid that is too small may miss fine structures—especially problematic when dealing with small or sparse objects—while a grid that is too large may introduce excessive background noise, reducing the precision of feature representation. To address this issue, we propose an enhanced PG-RCNN architecture with a Multi-Scale Grid Attention Module as the core contribution. This module improves the expressiveness of point features by aggregating multi-scale information and dynamically weighting features from different grid resolutions. Using a simple linear transformation, we generate attention weights to guide the model to focus on regions that contribute more to object recognition, while effectively filtering out redundant noise. We evaluate our method on the KITTI 3D object detection validation set. Experimental results show that, compared to the original PG-RCNN, our approach improves performance on the Cyclist category by 2.66% and 2.54% in the Moderate and Hard settings, respectively. Additionally, our approach shows more stable performance on small object detection tasks, with an average improvement of 2.57%, validating the positive impact of the Multi-Scale Grid Attention Module on fine-grained geometric modeling, and highlighting the efficiency and generalizability of our model. Full article

This paper presents a disaster detection framework based on aerial imagery, utilizing a Branch Convolutional Neural Network (B-CNN) to enhance feature learning efficiency. The B-CNN architecture incorporates branch training, enabling effective training and inference with reduced model parameters. To further optimize resource usage, the framework integrates DoReFa-Net for weight quantization and fixed-point parameter representation. An early exit mechanism is introduced to support low-latency, energy-efficient predictions. The proposed B-CNN hardware accelerator is implemented using TSMC 16 nm CMOS technology, incorporating power gating techniques to manage memory power consumption. Post-layout simulations demonstrate that the proposed hardware accelerator operates at 500 MHz with a power consumption of 37.56 mW. The system achieves a disaster prediction accuracy of 88.18%, highlighting its effectiveness and suitability for low-power, real-time applications in aerial disaster monitoring. Full article

Creating high-fidelity digital twins (DTs) for Industry 4.0 applications, it is fundamentally reliant on the accurate 3D modeling of physical assets, a task complicated by the inherent imperfections of real-world point cloud data. This paper addresses the challenge of reconstructing accurate, watertight, and topologically sound 3D meshes from sparse, noisy, and incomplete point clouds acquired in complex industrial environments. We introduce a robust two-stage completion-to-reconstruction framework, C2R3D-Net, that systematically tackles this problem. The methodology first employs a pretrained, self-supervised point cloud completion network to infer a dense and structurally coherent geometric representation from degraded inputs. Subsequently, a novel adaptive surface reconstruction network generates the final high-fidelity mesh. This network features a hybrid encoder (FKAConv-LSA-DC), which integrates fixed-kernel and deformable convolutions with local self-attention to robustly capture both coarse geometry and fine details, and a boundary-aware multi-head interpolation decoder, which explicitly models sharp edges and thin structures to preserve geometric fidelity. Comprehensive experiments on the large-scale synthetic ShapeNet benchmark demonstrate state-of-the-art performance across all standard metrics. Crucially, we validate the framework’s strong zero-shot generalization capability by deploying the model—trained exclusively on synthetic data—to reconstruct complex assets from a custom-collected industrial dataset without any additional fine-tuning. The results confirm the method’s suitability as a robust and scalable approach for 3D asset modeling, a critical enabling step for creating high-fidelity DTs in demanding, unseen industrial settings. Full article

Currently, quantum steganography schemes utilizing the least significant bit (LSB) approach are primarily optimized for fixed-point data processing, yet they encounter precision limitations when handling extended floating-point data structures owing to quantization error accumulation. To overcome precision constraints in quantum data hiding, the EPlsb-MFQS and MVlsb-MFQS quantum steganography algorithms are constructed based on the LSB approach in this study. The multichannel floating-point quantum representation of digital signals (MFQS) model enhances information hiding by augmenting the number of available channels, thereby increasing the embedding capacity of the LSB approach. Firstly, we analyze the limitations of fixed-point signals steganography schemes and propose the conventional quantum steganography scheme based on the LSB approach for the MFQS model, achieving enhanced embedding capacity. Moreover, the enhanced embedding efficiency of the EPlsb-MFQS algorithm primarily stems from the superposition probability adjustment of the LSB approach. Then, to prevent an unauthorized person easily extracting secret messages, we utilize channel qubits and position qubits as novel carriers during quantum message encoding. The secret message is encoded into the signal’s qubits of the transmission using a particular modulo value rather than through sequential embedding, thereby enhancing the security and reducing the time complexity in the MVlsb-MFQS algorithm. However, this algorithm in the spatial domain has low robustness and security. Therefore, an improved method of transferring the steganographic process to the quantum Fourier transformed domain to further enhance security is also proposed. This scheme establishes the essential building blocks for quantum signal processing, paving the way for advanced quantum algorithms. Compared with available quantum steganography schemes, the proposed steganography schemes achieve significant improvements in embedding efficiency and security. Finally, we theoretically delineate, in detail, the quantum circuit design and operation process. Full article