Search Results (8,815)

This paper proposes a novel dual-functional radar-communication (DFRC) framework that integrates unmanned aerial vehicle (UAV) communications into an integrated sensing and communication (ISAC) system, termed the ISAC-UAV architecture. In this system, the UAV’s mobility is leveraged to simultaneously serve multiple single-antenna uplink users (UEs) and perform radar-based sensing tasks. A key challenge stems from the target position uncertainty due to movement, which impairs matched filtering and beamforming, thereby degrading both uplink reception and sensing performance. Moreover, UAV energy consumption associated with mobility must be considered to ensure energy-efficient operation. We aim to jointly maximize radar sensing accuracy and minimize UAV movement energy over multiple time steps, while maintaining reliable uplink communications. To address this multi-objective optimization, we propose a deep reinforcement learning (DRL) framework based on a long short-term memory (LSTM)-enhanced deep deterministic policy gradient (DDPG) network. By leveraging historical target trajectory data, the model improves prediction of target positions, enhancing sensing accuracy. The proposed DRL-based approach enables joint optimization of UAV trajectory and uplink power control over time. Extensive simulations validate that our method significantly improves communication quality and sensing performance, while ensuring energy-efficient UAV operation. Comparative results further confirm the model’s adaptability and robustness in dynamic environments, outperforming existing UAV trajectory planning and resource allocation benchmarks. Full article

Considering the challenges posed by traditional continuous control set model predictive control (CCS-MPC) calculations, this paper proposes an event-triggered-based model predictive control (ET-MPC). First, a novel tracking error state-space model is proposed to improve tracking performance. Second, a reduced-order extended state observer (RESO) is designed to estimate and compensate for the total disturbances, thereby effectively improving robustness against the variations of the load resistance and reference voltage. At the same time, RESO significantly reduces computational complexity and accelerates the convergence speed of state estimation. Subsequently, an event trigger mechanism is introduced to enhance the MPC with a threshold function for the converter status. Finally, the reduced-order extended state observer-based model predictive control (RESO-MPC) is compared with the proposed ET-MPC through experiments. The ripple voltage of ET-MPC is within 2%, and the computational burden is reduced by more than 57%, verifying the effectiveness of the proposed ET-MPC. Full article

Multimodal sentiment analysis (MSA) faces key challenges such as incomplete modality inputs, long-range temporal dependencies, and suboptimal fusion strategies. To address these, we propose MGMR-Net, a Mamba-guided multimodal reconstruction and fusion network that integrates modality-aware reconstruction with text-centric fusion within an efficient state-space modeling framework. MGMR-Net consists of two core components: the Mamba-collaborative fusion module, which utilizes a two-stage selective state-space mechanism for fine-grained cross-modal alignment and hierarchical temporal integration, and the Mamba-enhanced reconstruction module, which employs continuous-time recurrence and dynamic gating to accurately recover corrupted or missing modality features. The entire network is jointly optimized via a unified multi-task loss, enabling simultaneous learning of discriminative features for sentiment prediction and reconstructive features for modality recovery. Extensive experiments on CMU-MOSI, CMU-MOSEI, and CH-SIMS datasets demonstrate that MGMR-Net consistently outperforms several baseline methods under both complete and missing modality settings, achieving superior accuracy, robustness, and generalization. Full article

This paper presents a comprehensive analysis comparing the implementation of electric and diesel buses for public transport services in the low-density area of the City of Jastrebarsko in Croatia. It utilizes a multidimensional approach and incorporates direct and indirect costs, such as vehicle acquisition, operation, charging, maintenance, and environmental impact costs during the lifecycle of the buses. The results show that, despite the higher initial investment in electric buses, these vehicles offer savings, especially when coupled with significantly reduced emissions of pollutants, which decreases indirect costs. However, local contexts differ, leading to a need to revise whether or not a municipality can finance the procurement and operations of such a fleet. The paper utilizes a robust methodological framework, integrating a proposal based on real-world data and demand and combining it with predictive analytics to forecast long-term benefits. The findings of the paper support the introduction of buses as a sustainable solution for Jastrebarsko, which provides insights for public transport planners, urban planners, and policymakers, with a discussion about the specific issues regarding the introduction, procurement, and operations of buses of different propulsion in a low-density area. Full article

In this study, a novel direct laser beam turning (DLBT) approach is proposed for the precision machining of AISI 308L austenitic stainless steel, which eliminates the need for cutting tools and thereby eradicates tool wear and vibration-induced surface irregularities. A nanosecond-pulsed Nd:YAG fiber laser (λ = 1064 nm, spot size = 0.05 mm) was used, and Ø1.6 mm × 20 mm cylindrical rods were processed under ambient conditions without auxiliary cooling. The experimental framework systematically evaluated the influence of scanning speed, pulse frequency, and the number of laser passes on dimensional accuracy and material removal efficiency. The results indicate that a maximum diameter reduction of 0.271 mm was achieved at a scanning speed of 3200 mm/s and 50 kHz, whereas 0.195 mm was attained at 6400 mm/s and 200 kHz. A robust second-order polynomial correlation (R² = 0.99) was established between diameter reduction and the number of passes, revealing the high predictability of the process. Crucially, when the scanning speed was doubled, the effective fluence was halved, considerably influencing the ablation characteristics. Despite the low fluence, evidence of material evaporation at elevated frequencies due to the incubation effect underscores the complex photothermal dynamics governing the process. This work constitutes the first comprehensive quantification of pass-dependent diameter modulation in DLBT and introduces a transformative, noncontact micromachining strategy for hard-to-machine alloys. The demonstrated precision, repeatability, and thermal control position DLBT as a promising candidate for next-generation manufacturing of high-performance miniaturized components. Full article

Landslide displacement prediction is crucial for disaster mitigation, yet traditional methods often fail to capture the complex, non-stationary spatiotemporal dynamics of slope evolution. This study introduces an enhanced prediction framework that integrates multi-scale signal processing with dynamic, geology-aware graph modeling. The proposed methodology first employs the Maximum Overlap Discrete Wavelet Transform (MODWT) to denoise raw Global Navigation Satellite System (GNSS)-monitored displacement time series data, enhancing the underlying deformation features. Subsequently, a geology-aware graph is constructed, using the temporal correlation of displacement series as a practical proxy for physical relatedness between monitoring nodes. The framework’s core innovation lies in a dynamic graph optimization model with low-rank constraints, which adaptively refines the graph topology to reflect time-varying inter-sensor dependencies driven by factors like mining activities. Experiments conducted on a real-world dataset from an active open-pit mine demonstrate the framework’s superior performance. The DCRNN-proposed model achieved the highest accuracy among eight competing models, recording a Root Mean Square Error (RMSE) of 2.773 mm in the Vertical direction, a 39.1% reduction compared to its baseline. This study validates that the proposed dynamic graph optimization approach provides a robust and significantly more accurate solution for landslide prediction in complex, real-world engineering environments. Full article

Financial market forecasting remains challenging due to complex nonlinear dynamics and regime-dependent behaviors that traditional models struggle to capture effectively. This research introduces the Adaptive Financial Reservoir Network with Hypernetwork Flow (AFRN–HyperFlow) framework, a novel ensemble architecture integrating Echo State Networks, temporal convolutional networks, mixture density networks, adaptive Hypernetworks, and deep state-space models for enhanced financial time-series prediction. Through comprehensive feature engineering incorporating technical indicators, spectral decomposition, reservoir-based representations, and flow dynamics characteristics, the framework achieves superior forecasting performance across diverse market conditions. Experimental validation on 26,817 balanced samples demonstrates exceptional results with an F1-score of 0.8947, representing a 12.3% improvement over State-of-the-Art baseline methods, while maintaining robust performance across asset classes from equities to cryptocurrencies. The adaptive Hypernetwork mechanism enables real-time regime-change detection with 2.3 days average lag and 95% accuracy, while systematic SHAP analysis provides comprehensive interpretability essential for regulatory compliance. Ablation studies reveal Echo State Networks contribute 9.47% performance improvement, validating the architectural design. The AFRN–HyperFlow framework addresses critical limitations in uncertainty quantification, regime adaptability, and interpretability, offering promising directions for next-generation financial forecasting systems incorporating quantum computing and federated learning approaches. Full article

Understanding alpine plants’ survival and reproduction is crucial for their conservation in climate change. Based on 423 valid distribution points, this study utilizes the MaxEnt model to predict the potential habitat and distribution dynamics of Leontopodium nanum under both current and future climate scenarios, while clarifying the key factors that influence its distribution. The primary ecological drivers of distribution are altitude (2886.08 m–5576.14 m) and the mean temperature of the driest quarter (−6.60–1.55 °C). Currently, the suitable habitat area is approximately 520.28 × 10⁴ km², covering about 3.5% of the global land area, concentrated mainly in the Tibetan Plateau, with smaller regions across East and South Asia. Under future climate scenarios, low-emission (SSP126), suitable areas are projected to expand during the 2050s and 2070s. High-emission (SSP585), suitable areas may decrease by 50%, with a 66.07% reduction in highly suitable areas by the 2070s. The greatest losses are expected in the south-eastern Tibetan Plateau. Regarding dynamic habitat changes, by the 2050s, newly suitable areas will account for 51.09% of the current habitat, while 68.26% of existing habitat will become unsuitable. By the 2070s, newly suitable areas will rise to 71.86% of the current total, but the loss of existing areas will exceed these gains, particularly under the high-emission scenario. The centroid of suitable habitats is expected to shift northward, with migration distances ranging from 23.94 km to 342.42 km. The most significant shift is anticipated under the SSP126 scenario by the 2070s. This study offers valuable insights into the distribution dynamics of L. nanum and other alpine species under the context of climate change. From a conservation perspective, it is recommended to prioritize the protection and restoration of vegetation in key habitat patches or potential migration corridors, restrict overgrazing and infrastructure development, and maintain genetic diversity and dispersal capacity through assisted migration and population genetic monitoring when necessary. These measures aim to provide a robust scientific foundation for the comprehensive conservation and sustainable management of the grassland ecosystem on the Qinghai–Tibet Plateau. Full article

In today’s dynamic and highly uncertain environment, organizations, particularly in the aviation sector, face increasing challenges that demand resilient, flexible, and data-driven change management decisions. Responding to the growing need for structured approaches to managing complex uncertainties—geopolitical tensions, economic volatility, social shifts, rapid technological advancements, environmental pressures and regulatory changes—this research proposes a theoretical change management model for aviation service providers, such as airports. Integrating three analytical approaches, the model offers a robust, multi-method approach for supporting sustainable transformation under uncertainty. Normative analysis using Bayesian decision theory identifies influential external environmental factors, capturing probabilistic relationships, and revealing causal links under uncertainty. Prescriptive planning through scenario theory explores alternative future pathways and helps to identify possible predictions, offer descriptive evaluation employing fuzzy comprehensive evaluation, and assess decision quality under vagueness and complexity. The proposed four-stage model—observation, analysis, evaluation, and response—offers a methodology for continuous external environment monitoring, scenario development, and data-driven, proactive change management decision-making, including the impact assessment of change and development. The proposed model contributes to the theoretical advancement of the change management research area under uncertainty and offers practical guidance for aviation organizations (airports) facing a volatile external environment. This framework strengthens aviation organizations’ ability to anticipate, evaluate, and adapt to multifaceted external changes, supporting operational flexibility and adaptability and contributing to the sustainable development of aviation services. Supporting aviation organizations with tools to proactively manage systemic uncertainty, this research directly supports the integration of sustainability principles, such as resilience and adaptability, for long-term value creation through change management decision-making. Full article

Many commercial wearable sensor systems typically rely on a single continuous cardiorespiratory sensing modality, photoplethysmography (PPG), which suffers from inherent biases (i.e., differences in skin tone) and noise (e.g., motion and pressure artifacts). In this research, we present a wearable device that provides robust estimates of cardiorespiratory variables by combining three physiological signals from the upper arm: multiwavelength PPG, single-sided electrocardiography (SS-ECG), and bioimpedance plethysmography (BioZ), along with an inertial measurement unit (IMU) providing 3-axis accelerometry and gyroscope information. We evaluated the multimodal device on 16 subjects by its ability to estimate heart rate (HR) and breathing rate (BR) in the presence of various static and dynamic noise sources (e.g., skin tone and motion). We proposed a hierarchical approach that considers the subject’s skin tone and signal quality to select the optimal sensing modality for estimating HR and BR. Our results indicate that, when estimating HR, there is a trade-off between accuracy and robustness, with SS-ECG providing the highest accuracy (low mean absolute error; MAE) but low reliability (higher rates of sensor failure), and PPG/BioZ having lower accuracy but higher reliability. When estimating BR, we find that fusing estimates from multiple modalities via ensemble bagged tree regression outperforms single-modality estimates. These results indicate that multimodal approaches to cardiorespiratory monitoring can overcome the accuracy–robustness trade-off that occurs when using single-modality approaches. Full article

The accurate estimation of urban property values is a key challenge for appraisers, market participants, financial institutions, and urban planners. In recent years, machine learning (ML) techniques have emerged as promising tools for price forecasting due to their ability to model complex relationships among variables. However, their application raises two main critical issues: (i) the risk of overfitting, especially with small datasets or with noisy data; (ii) the interpretive issues associated with the “black box” nature of many models. Within this framework, this paper proposes a methodological approach that addresses both these issues, comparing the predictive performance of three ML algorithms—k-Nearest Neighbors (kNN), Random Forest (RF), and the Artificial Neural Network (ANN)—applied to the housing market in the city of Salerno, Italy. For each model, overfitting is preliminarily assessed to ensure predictive robustness. Subsequently, the results are interpreted using explainability techniques, such as SHapley Additive exPlanations (SHAPs) and Permutation Feature Importance (PFI). This analysis reveals that the Random Forest offers the best balance between predictive accuracy and transparency, with features such as area and proximity to the train station identified as the main drivers of property prices. kNN and the ANN are viable alternatives that are particularly robust in terms of generalization. The results demonstrate how the defined methodological framework successfully balances predictive effectiveness and interpretability, supporting the informed and transparent use of ML in real estate valuation. Full article

The COVID-19 pandemic catalyzed the rapid adoption of remote patient monitoring (RPM) technologies such as telemedicine and wearable devices (WDs), significantly transforming healthcare delivery. Telemedicine made virtual consultations possible, reducing in-person visits and infection risks, particularly for the management of chronic diseases. Wearable devices enabled the real-time continuous monitoring of health that assisted in condition prediction and management, such as for COVID-19. This narrative review addresses these transformations by uniquely synthesizing findings from 13 diverse studies (sourced from PubMed and Google Scholar, 2020–2024) to analyze the parallel evolution of telemedicine and WDs as interconnected RPM components. It highlights the pandemic’s dual impact, as follows: accelerating RPM innovation and adoption while simultaneously unmasking systemic challenges such as inequities in access and a need for robust integration approaches; while telemedicine usage soared during the pandemic, consumption post-pandemic, as indicated by the reviewed studies, suggests continued barriers to adoption among older adults. Likewise, wearable devices demonstrated significant potential in early disease detection and long-term health management, with promising applications extending beyond COVID-19, including long COVID conditions. Addressing the identified challenges is crucial for healthcare providers and systems to fully embrace these technologies and this would improve efficiency and patient outcomes. Full article

Accurate prediction of the coal abrasive index (AI) is critical for optimizing coal processing efficiency and minimizing equipment wear in industrial applications. This study explores tree-based machine learning models; Random Forest (RF), Gradient Boosting Trees (GBT), and Extreme Gradient Boosting (XGBoost) to predict AI using selected coal properties. A database of 112 coal samples from the KwaZulu-Natal Coalfield in South Africa was used. Initial predictions using all eight input properties revealed suboptimal testing performance (R²: 0.63–0.72), attributed to outliers and noisy data. Feature importance analysis identified calorific value, quartz, ash, and Pyrite as dominant predictors, aligning with their physicochemical roles in abrasiveness. After data cleaning and feature selection, XGBoost achieved superior accuracy (R² = 0.92), outperforming RF (R² = 0.85) and GBT (R² = 0.81). The results highlight XGBoost’s robustness in modeling non-linear relationships between coal properties and AI. This approach offers a cost-effective alternative to traditional laboratory methods, enabling industries to optimize coal selection, reduce maintenance costs, and enhance operational sustainability through data-driven decision-making. Additionally, quartz and Ash content were identified as the most influential parameters on AI using the Cosine Amplitude technique, while calorific value had the least impact among the selected features. Full article

This paper introduces a knowledge–data dual-driven method for predicting groundwater conditions during tunnel construction. Unlike existing methods, our approach effectively integrates trend characteristics of apparent resistivity from detection results with geological distribution characteristics and expert insights. This dual-driven strategy significantly enhances the accuracy of the prediction model. The intelligent prediction process for tunnel groundwater conditions proceeds in the following steps: First, the apparent resistivity data matrix is obtained from transient electromagnetic detection results and standardized. Second, to improve data quality, trend characteristics are extracted from the apparent resistivity data, and outliers are eliminated. Third, expert insights are systematically integrated to fully utilize prior information on groundwater conditions at the construction face, leading to the establishment of robust predictive models tailored to data from various construction surfaces. Finally, the relevant prediction segment is extracted to complete the groundwater condition forecast. Full article

Rotating compressors are foundational in various industrial processes, particularly in the oil-and-gas sector, where reliable fault detection is crucial for maintaining operational continuity. While Graph Attention Network (GAT) frameworks are widely available, this study advances the state of the art by introducing a Bayesian GAT method specifically tailored for vibration-based compressor fault diagnosis. The approach integrates domain-specific digital-twin simulations built with Rotordynamic software (1.3.0), and constructs dual adjacency matrices to encode both physically informed and data-driven sensor relationships. Additionally, a hybrid forecasting-and-reconstruction objective enables the model to capture short-term deviations as well as long-term waveform fidelity. Monte Carlo dropout further decomposes prediction uncertainty into aleatoric and epistemic components, providing a more robust and interpretable model. Comparative evaluations against conventional Long Short-Term Memory (LSTM)-based autoencoder and forecasting methods demonstrate that the proposed framework achieves superior fault-detection performance across multiple fault types, including misalignment, bearing failure, and unbalance. Moreover, uncertainty analyses confirm that fault severity correlates with increasing levels of both aleatoric and epistemic uncertainty, reflecting heightened noise and reduced model confidence under more severe conditions. By enhancing GAT fundamentals with a domain-tailored dual-graph strategy, specialized Bayesian inference, and digital-twin data generation, this research delivers a comprehensive and interpretable solution for compressor fault diagnosis, paving the way for more reliable and risk-aware predictive maintenance in complex rotating machinery. Full article