Search Results (562)

Existing nighttime visibility enhancement methods primarily focus on improving overall brightness under low-light conditions. However, nighttime road images are also affected by glare, glow, and flare from complex light sources such as streetlights and headlights, making it challenging to suppress locally overexposed regions and recover fine details. To address these challenges, we propose a Nighttime Road Glare Suppression Network (NRGS-Net) for glare removal and detail restoration. Specifically, to handle diverse glare disturbances caused by the uncertainty in light source positions and shapes, we designed a gated positional attention (GPA) module that integrates positional encoding with local contextual information to guide the network in accurately locating and suppressing glare regions, thereby enhancing the visibility of affected areas. Furthermore, we introduced an improved Uformer backbone named LCAtransformer, in which the downsampling layers adopt efficient depthwise separable convolutions to reduce computational cost while preserving critical spatial information. The upsampling layers incorporate a residual PixelShuffle module to achieve effective restoration in glare-affected regions. Additionally, channel attention is introduced within the Local Context-Aware Feed-Forward Network (LCA-FFN) to enable adaptive adjustment of feature weights, effectively suppressing irrelevant and interfering features. To advance the research in nighttime glare suppression, we constructed and publicly released the Night Road Glare Dataset (NRGD) captured in real nighttime road scenarios, enriching the evaluation system for this task. Experiments conducted on the Flare7K++ and NRGD, using five evaluation metrics and comparing six state-of-the-art methods, demonstrate that our method achieves superior performance in both subjective and objective metrics compared to existing advanced methods. Full article

This paper studies deep reinforcement learning (DRL)-based joint resource allocation and three-dimensional (3D) trajectory optimization for unmanned aerial vehicle (UAV)–ground access point (GAP) cooperative non-orthogonal multiple access (NOMA) communication in Industrial Internet of Things (IIoT) systems. Cooperative and non-cooperative users adopt different signal transmission strategies to meet diverse, task-oriented, quality-of-service requirements. Specifically, the DRL framework based on the Soft Actor–Critic algorithm is proposed to jointly optimize user scheduling, power allocation, and UAV trajectory in continuous action spaces. Closed-form power allocation and maximum weight bipartite matching are integrated to enable efficient user pairing and resource management. Simulation results show that the proposed scheme significantly enhances system performance in terms of throughput, spectral efficiency, and interference management, while enabling robustness against channel uncertainties in dynamic IIoT environments. The findings indicate that combining model-free reinforcement learning with conventional optimization provides a viable solution for adaptive resource management in dynamic UAV-GAP cooperative communication scenarios. Full article

In this study, an accurate attenuation measurement system with high attenuation capability (≥100 dB) is presented, covering a broad radio frequency range from 1 GHz to 25 GHz. The system employs a dual-channel intermediate frequency (IF) substitution method, utilizing a programmable inductive voltage divider (IVD) that provides precise voltage ratios at a 1 kHz operating IF, serving as the primary attenuation standard. To ensure optimal inter-channel isolation, essential for accurate high-attenuation measurements, an optical fiber assembly, consisting of a laser diode, a wideband external electro-optic modulator, and a photodetector, is integrated between the channels. A comprehensive performance evaluation is presented, with particular emphasis on the programmable IVD calibration technique, which achieves an accuracy better than 0.001 dB across all attenuation levels, and on the role of the optical fiber assembly in enhancing isolation, demonstrating levels exceeding 120 dB across the entire frequency range. The system demonstrates measurement capabilities with expanded uncertainties (k = 2) of 0.004 dB, 0.008 dB, and 0.010 dB at attenuation levels of 20 dB, 60 dB, and 100 dB, respectively. Full article

This study investigates the downstream wake characteristics of a helical hydrokinetic turbine through combined experimental and numerical analyses. A four-bladed helical turbine with a 20 cm rotor diameter and blockage ratio of 53.57% was tested in an open water channel under a flow rate of 180 m³/h, corresponding to a Reynolds number of approximately 90 × 10³. Velocity measurements were collected at 13 downstream cross-sections using an Acoustic Doppler Velocimeter, with each point sampled repeatedly. Standard error analysis was applied to quantify measurement uncertainty. Complementary numerical simulations were conducted in ANSYS Fluent using a steady-state k-ω Shear Stress Transport (SST) turbulence model, with a mesh of 4.7 million elements and mesh independence confirmed. Velocity deficit and turbulence intensity were employed as primary parameters to characterize the wake structure, while the analysis also focused on the recovery of cross-sectional velocity profiles to validate the extent of wake influence. Experimental results revealed a maximum velocity deficit of over 40% in the near-wake region, which gradually decreased with downstream distance, while turbulence intensity exceeded 50% near the rotor and dropped below 10% beyond 4 m. In comparison, numerical findings showed a similar trend but with lower peak velocity deficits of 16.6%. The root mean square error (RMSE) and mean absolute error (MAE) between experimental and numerical mean velocity profiles were calculated as 0.04486 and 0.03241, respectively, demonstrating reasonable agreement between the datasets. Extended simulations up to 30 m indicated that flow profiles began to resemble ambient conditions around 18–20 m. The findings highlight the importance of accurately identifying the downstream distance at which the wake effect fully dissipates, as this is crucial for determining appropriate inter-turbine spacing. The study also discusses potential sources of discrepancies between experimental and numerical results, as well as the limitations of the modeling approach. Full article

With the growing global demand for clean energy, the accuracy of wind-power forecasting plays a vital role in ensuring the stable operation of power systems. However, wind-power generation is significantly influenced by meteorological conditions and is characterized by high uncertainty and multiscale fluctuations. Traditional recurrent neural network (RNN) and long short-term memory (LSTM) models, although capable of handling sequential data, struggle with modeling long-term temporal dependencies due to the vanishing gradient problem; thus, they are now rarely used. Recently, Transformer models have made notable progress in sequence modeling compared to RNNs and LSTM models. Nevertheless, when dealing with long wind-power sequences, their quadratic computational complexity (O(L²)) leads to low efficiency, and their global attention mechanism often fails to capture local periodic features accurately, tending to overemphasize redundant information while overlooking key temporal patterns. To address these challenges, this paper proposes a wind-power forecasting method based on dimension-transformed collaborative multidimensional multiscale attention (DTCMMA). This method first employs fast Fourier transform (FFT) to automatically identify the main periodic components in wind-power data, reconstructing the one-dimensional time series as a two-dimensional spatiotemporal representation, thereby explicitly encoding periodic features. Based on this, a collaborative multidimensional multiscale attention (CMMA) mechanism is designed, which hierarchically integrates channel, spatial, and pixel attention to adaptively capture complex spatiotemporal dependencies. Considering the geometric characteristics of the reconstructed data, asymmetric convolution kernels are adopted to enhance feature extraction efficiency. Experiments on multiple wind-farm datasets and energy-related datasets demonstrate that DTCMMA outperforms mainstream methods such as Transformer, iTransformer, and TimeMixer in long-sequence forecasting tasks, achieving improvements in MSE performance by 34.22%, 2.57%, and 0.51%, respectively. The model’s training speed also surpasses that of the fastest baseline by 300%, significantly improving both prediction accuracy and computational efficiency. This provides an efficient and accurate solution for wind-power forecasting and contributes to the further development and application of wind energy in the global energy mix. Full article

Motor Imagery (MI) classification plays a crucial role in enhancing the performance of brain–computer interface (BCI) systems, thereby enabling advanced neurorehabilitation and the development of intuitive brain-controlled technologies. However, MI classification using electroencephalography (EEG) is hindered by spatiotemporal variability and the limited interpretability of deep learning (DL) models. To mitigate these challenges, dropout techniques are employed as regularization strategies. Nevertheless, the removal of critical EEG channels, particularly those from the sensorimotor cortex, can result in substantial spatial information loss, especially under limited training data conditions. This issue, compounded by high EEG variability in subjects with poor performance, hinders generalization and reduces the interpretability and clinical trust in MI-based BCI systems. This study proposes a novel framework integrating channel dropout—a variant of Monte Carlo dropout (MCD)—with class activation maps (CAMs) to enhance robustness and interpretability in MI classification. This integration represents a significant step forward by offering, for the first time, a dedicated solution to concurrently mitigate spatiotemporal uncertainty and provide fine-grained neurophysiologically relevant interpretability in motor imagery classification, particularly demonstrating refined spatial attention in challenging low-performing subjects. We evaluate three DL architectures (ShallowConvNet, EEGNet, TCNet Fusion) on a 52-subject MI-EEG dataset, applying channel dropout to simulate structural variability and LayerCAM to visualize spatiotemporal patterns. Results demonstrate that among the three evaluated deep learning models for MI-EEG classification, TCNet Fusion achieved the highest peak accuracy of 74.4% using 32 EEG channels. At the same time, ShallowConvNet recorded the lowest peak at 72.7%, indicating TCNet Fusion’s robustness in moderate-density montages. Incorporating MCD notably improved model consistency and classification accuracy, especially in low-performing subjects where baseline accuracies were below 70%; EEGNet and TCNet Fusion showed accuracy improvements of up to 10% compared to their non-MCD versions. Furthermore, LayerCAM visualizations enhanced with MCD transformed diffuse spatial activation patterns into more focused and interpretable topographies, aligning more closely with known motor-related brain regions and thereby boosting both interpretability and classification reliability across varying subject performance levels. Our approach offers a unified solution for uncertainty-aware, and interpretable MI classification. Full article

The unique vegetation in riparian zones is fundamental for various ecological and socio-economic functions in these transitional areas. Sustainable management requires detailed spatial information about the occurring flora. Here, we present a Deep Learning (DL)-based approach for processing multimodal high-resolution remote sensing data (aerial RGB and near-infrared (NIR) images and elevation maps) to generate a classification map of the tidal Elbe and a section of the Rhine River (Germany). The ground truth was based on existing mappings of vegetation and biotope types. The results showed that (I) despite a large class imbalance, for the tidal Elbe, a high mean Intersection over Union (IoU) of about 78% was reached. (II) At the Rhine River, a lower mean IoU was reached due to the limited amount of training data and labelling errors. Applying transfer learning methods and labelling error correction increased the mean IoU to about 60%. (III) Early fusion of the modalities was beneficial. (IV) The performance benefits from using elevation maps and the NIR channel in addition to RGB images. (V) Model uncertainty was successfully calibrated by using temperature scaling. The generalization ability of the trained model can be improved by adding more data from future aerial surveys. Full article

Climate change risk has significant impacts on corporate financial activities. Using firm-level data from A-share listed companies in China from 2010 to 2022, we examine how climate risk affects corporate debt financing capacity. We find that climate change risk significantly weakens firms’ ability to raise debt, leading to lower leverage and higher financing costs. These results remain robust across various checks for endogeneity and alternative specifications. We also show that reducing corporate carbon emission intensity can mitigate the negative impact of climate risk on debt financing, suggesting that supply-side credit policies are more effective than demand-side capital structure choices. Furthermore, we identify three channels through which climate risk impairs debt capacity: reduced competitiveness, increased default risk, and diminished resilience. Our heterogeneity analysis reveals that these adverse effects are more pronounced for non-state-owned firms, firms with weaker internal controls, and companies in highly financialized regions, and during periods of heightened environmental uncertainty. We also apply textual analysis and machine learning to the measurement of climate change risks, partially mitigating the geographic biases and single-dimensional shortcomings inherent in macro-level indicators, thus enriching the quantitative research on climate change risks. These findings provide valuable insights for policymakers and financial institutions in promoting corporate green transition, guiding capital allocation, and supporting sustainable development. Full article

Thermoacoustic devices, such as refrigerators and heat pumps, present unique measurement challenges due to the simultaneous presence of rapidly fluctuating acoustic parameters and more stable thermal variables. Accurate and informative measurements during operation are crucial for developing effective control algorithms and optimizing performance under specific conditions. However, issues like inappropriate sampling frequencies and excessive data storage can lead to unintended averaging, compromising measurement quality. This study introduces a comprehensive experimental procedure aimed at enhancing the reliability of measurements in thermoacoustic systems. The approach encompasses meticulous experimental design, identification of measurement uncertainties and influencing factors during standard operation, and a statistical uncertainty analysis. Experimental findings reveal a significant reduction in temperature measurement uncertainty with increased thermoacoustic channel length and highlight the substantial impact of device structural features on performance. These insights are instrumental for refining measurement protocols and advancing the development of efficient thermoacoustic technologies. Full article

In DC microgrids, optimizing the hybrid energy storage system (HESS) current control to meet the power requirements of the load is generally a difficult and challenging task. This is because the HESS always operates under various load conditions, which are influenced by measurement disturbances and parameter uncertainties. Therefore, in this paper, we propose the H_∞ state feedback control based on the reference modulation to improve the current tracking errors of the battery (Bat) and supercapacitor (SC) in the HESS for power tracking performance. Without altering the system control signal, the reference modulation technique combines the feedforward channel and output feedback signal directly to modulate the required currents of the Bat and SC derived from the required load power. The H_∞ state feedback control based on the required Bat and SC currents modulated by the reference modulation technique is proposed to improve the current tracking errors under the influence of measurement disturbances and parameter uncertainties without a disturbance observer. The ability of the reference modulation technique to attenuate the disturbance without the use of a disturbance observer is one advantage for improving transient performance. The improvement of the HESS’s power tracking performance in DC microgrids is confirmed by study results presented under the influence of measurement disturbances for nominal parameters and parameter uncertainties. Full article

Clamp-on ultrasonic flowmeters serve as an important tool for on-site testing of gas flow meters. However, its accuracy is significantly affected by the actual flow field, thus limiting its application scenarios. To address this issue, this study focuses on typical industrial disturbance structures and obtains the evolution and distribution of non-ideal flow fields downstream of disturbances through experiments and numerical simulations, as well as their effects on velocity and flow measurement errors. The results indicate that when traditional reflection or diagonal measurements were used in the downstream of disturbances, the flow deviation was largely dependent on the installation position and angle of the clamp-on ultrasonic flowmeter. This introduced significant uncertainty and bias, rendering it impossible to correct measurement results through quantitative coefficients. Utilizing a dual-channel measurement method can enhance measurement accuracy. When two sets of sensors perpendicular to each other were used to combine the reflection measurement path, the deviation fluctuation downstream of disturbances can be effectively controlled within the range of ±2%, irrespective of the installation angle. This measurement approach significantly reduced the distance limitations on the distance of the straight pipe section during the use of clamp-on ultrasonic flowmeters. Full article

The conventional practice in the design of storm drainage systems is based on statistically stationary load and resistance conditions that remain invariant over time. However, uncertainties in the variables affect the design accuracy and the satisfactory performance of these hydrosystems during their operation and service. To overcome this limitation, a design methodology for a storm drainage channel was proposed using a probabilistic framework that incorporates uncertainty analysis of random variables and estimates the system’s probability of failure in terms of design depth and maximum allowable velocity. This methodology employs the Monte Carlo simulation technique and offers an alternative design and analysis approach to strengthen the conventional sizing method for drainage channels in urban watersheds. Based on uncertainty criteria associated with hydraulic design, operation, and prospective changes in the watershed and the channel, appropriate dimensions were estimated regarding design depth and freeboard. The results of this study demonstrate that the annual probability of failure of a channel, when considering uncertainty, is significantly higher than the yearly exceedance probability associated with the hydrological design return period event. Therefore, the proposed methodology is appropriate for estimating the system’s capacity and potential failure risk. This methodology may also be applied to sizing other stormwater drainage structures. Full article

Probabilistic selling is a retail strategy in which consumers purchase products without knowing their exact identities until after purchase, with various applications like gaming and retail; a real-world practice involves retailers may unsealing and reselling goods to meet consumer demand for transparency. This disrupts manufacturers’ strategies designed to adopt the uncertainty for segmentation and pricing. Using a vertically differentiated supply chain model structured as a Stackelberg game framework, this study examines how transparency from retailer unsealing affects profitability, consumer surplus, and market dynamics. Key findings include the following: (1) Unsealing increases retailer profits by aligning pricing with heterogeneous consumer willingness to pay. (2) Introducing a manufacturer’s direct channel reduces unsealing profits via price competition. (3) Unsealing creates conflicts between manufacturers’ design goals and retailers’ profit-driven incentives. By applying a Stackelberg game framework to model unsealing as a downstream transparency decision, this work advances the probabilistic selling literature by offering a structured approach to analyzing how downstream transparency and retailer strategies reshape probabilistic selling and supply chain dynamics. It highlights the need for manufacturers to balance segmentation, pricing, and channel control, offering insights into mitigating conflicts between design intentions and downstream market behaviors. Full article

Using 2010–2019 data from 404 listed construction companies in China, we explore the relationship between perception of economic policy uncertainty (PEPU) and energy consumption intensity (ECI) based on a fixed effects model controlling for company, year, and city fixed effects, with standard errors clustered at the industry level. The results show that the perception of economic policy uncertainty reduces construction enterprise energy consumption intensity, and this result holds after a series of robustness and endogeneity tests. Further, this effect is stronger in firms with more green shareholders, environmental information disclosure, and external attention. Moreover, mechanism analysis indicates that internal control enhancement and green innovation improvement, including quantity and quality, are the underlying channels through which the perception of economic policy uncertainty influences energy consumption intensity. Full article

Rip currents at featureless beaches (i.e., beaches lacking sandbars or channels) are often hydrodynamically controlled, exhibiting intermittent and unpredictable behaviors that pose significant risks to recreational beach users. This study assessed occurrences of rip currents under a range of idealized morphology configurations and hydrodynamic wave forcing parameters using a wave-resolving Boussinesq-type model. Numerical experiments revealed that rip currents with durations on the time scale of 10 min are generated in the forms of vortex pairs, intensified eddies, mega-rips, and eddies shedding from longshore currents. In general, the key conditions that promote rip current formation at featureless beaches include shoreline curvature, headlands, moderately mild beach slopes (e.g., 0.02–0.03), normal or near-normal wave incidence, and large wave heights. Most importantly, this study highlights inherent uncertainties in rip current occurrences, particularly under conditions usually perceived as low risk: low wave heights, short wave periods, oblique wave incidence, and straight shorelines. These conditions can lead to transient rip currents and pose an unexpected hazard that coastal communities should be aware of. Full article