Search Results (105)

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

Automated defect detection for printed fabric manufacturing faces critical challenges in balancing industrial-grade accuracy with real-time deployment efficiency. To address this, we propose RT-DETR-FFD, a knowledge-distilled detector optimized for printed fabric defect inspection. Firstly, the student model integrates a Fourier cross-stage mixer (FCSM). This module disentangles defect features from periodic textile backgrounds through spectral decoupling. Secondly, we introduce FuseFlow-Net to enable dynamic multi-scale interaction, thereby enhancing discriminative feature representation. Additionally, a learnable positional encoding (LPE) module transcends rigid geometric constraints, strengthening contextual awareness. Furthermore, we design a dynamic correlation-guided loss (DCGLoss) for distillation optimization. Our loss leverages masked frequency-channel alignment and cross-domain fusion mechanisms to streamline knowledge transfer. Experiments demonstrate that the distilled model achieves an mAP@0.5 of 82.1%, surpassing the baseline RT-DETR-R18 by 6.3% while reducing parameters by 11.7%. This work establishes an effective paradigm for deploying high-precision defect detectors in resource-constrained industrial scenarios, advancing real-time quality control in textile manufacturing. Full article

In this study, the treatment of pharmaceutical tail water (PTW) by coagulation, Fenton combined with membrane bioreactor (MBR), was studied. Optimal parameters were obtained according to batch experiment and central composite design (CCD). Results showed that Polymeric Ferric Sulfate (PFS) was the best coagulant for original pharmaceutical tailwater due to less dosage and higher removal efficiency to TOC, COD, NH₄⁺-N and UV_254m, with the optimized pH = 7.25 and 0.53 g/L PFS dosage. The best coagulation performance was achieved when the mixer was stirred at 250 rpm for 3 min, 60 rpm for 10 min, and then left to stand for 60 min. Coagulation mainly removed organics with molecular weight above 10 kDa. After treated by coagulation, 43.1% TOC removal efficiency of PTW was obtained by Fenton reaction with 11.6 mmol/L H₂O₂, 3.0 mmol/L FeSO₄, pH = 3.3 and T = 50 min. A type of common macromolecule aromatic amino acid compounds which located Ex = 250 nm and Em = 500 nm was the main reason that caused the high TOC concentration in the effluent. Stable COD and NH₄⁺-N removal efficiencies in the MBR reactor within 10 d were observed when the mixture of pre-treated PTW (20%, v) and domestic sewage (80%, v) was fed into the MBR reactor, and over 95% COD and 50% NH₄⁺-N were removed. One kind of amino acid similar to tryptophan was the prime reason that caused PTW resistance to be degraded. Analysis of the microorganism community in the MBR suggested that norank_f__Saprospiraceae was the key microorganism in degrading of PTW. Full article

Remote sensing target detection faces significant challenges in handling multi-scale targets, with the high similarity in color and shape between targets and backgrounds in complex scenes further complicating the detection task. To address this challenge, we propose a multi-Scale and complex Background Aware network for remote sensing target detection, named YOLO-SBA. Our proposed YOLO-SBA first processes the input through the Multi-Branch Attention Feature Fusion Module (MBAFF) to extract global contextual dependencies and local detail features. It then integrates these features using the Bilateral Attention Feature Mixer (BAFM) for efficient fusion, enhancing the saliency of multi-scale target features to tackle target scale variations. Next, we utilize the Gated Multi-scale Attention Pyramid (GMAP) to perform channel–spatial dual reconstruction and gating fusion encoding on multi-scale feature maps. This enhances target features while finely suppressing spectral redundancy. Additionally, to prevent the loss of effective information extracted by key modules during inference, we improve the downsampling method using Asymmetric Dynamic Downsampling (ADDown), maximizing the retention of image detail information. We achieve the best performance on the DIOR, DOTA, and RSOD datasets. On the DIOR dataset, YOLO-SBA improves mAP by 16.6% and single-category detection AP by 0.8–23.8% compared to the existing state-of-the-art algorithm. Full article

Precise identification of water bodies in agricultural watersheds is crucial for irrigation, water resource management, and flood disaster prevention. However, the spectral noise caused by complex light and shadow interference and water quality differences, combined with the diverse shapes of water bodies and the high computational cost of image processing, severely limits the accuracy of water body recognition in agricultural watersheds. This paper proposed a lightweight and efficient learnable Kalman filter and Deformable Convolutional Attention Network (LKF-DCANet). The encoder is built using a shallow Channel Attention-Enhanced Deformable Convolution module (CADCN), while the decoder combines a Convolutional Additive Token Mixer (CATM) and a learnable Kalman filter (LKF) to achieve adaptive noise suppression and enhance global context modeling. Additionally, a feature-based knowledge distillation strategy is employed to further improve the representational capacity of the lightweight model. Experimental results show that LKF-DCANet achieves an Intersection over Union (IoU) of 85.95% with only 0.22 M parameters on a public dataset. When transferred to a self-constructed UAV dataset, it achieves an IoU of 96.28%, demonstrating strong generalization ability. All experiments are conducted on RGB optical imagery, confirming that LKF-DCANet offers an efficient and highly versatile solution for water body segmentation in precision agriculture. Full article

Static mixers have been widely used in marine research fields, such as marine control systems, ballast water treatment systems, and seawater desalination, due to their high efficiency, low energy consumption, and broad applicability. However, the turbulent mixing process and fluid–wall interactions involving complex structures make the mixing transport characteristics of static mixers complex and nonlinear, which affect the mixing efficiency and stability of the fluid control device. Here, the modeling and design optimization of the two-phase flow mixing and transport dynamics of a static mixer face many challenges. This paper proposes a modeling and problem-solving method for the two-phase flow transport dynamics of static mixers, based on the lattice Boltzmann method (LBM) and large eddy simulation (LES). The characteristics of the two-phase flow mixing dynamics and design optimization strategies for complex component structures are analyzed. First, a two-phase flow transport dynamics model for static mixers is set up, based on the LBM and a multiple-relaxation-time wall-adapting local eddy (MRT-WALE) vortex viscosity coupling model. Using octree lattice block refinement technology, the interaction mechanism between the fluid and the wall during the mixing process is explored. Then, the design optimization strategies for the flow field are analyzed under different flow rates and mixing element configurations to improve the mixing efficiency and stability. The research results indicate that the proposed modeling and problem-solving methods can reveal the dynamic evolution process of mixed-flow fields. Blade components are the main driving force behind the increased turbulent kinetic energy and induced vortex formation, enhancing the macroscopic mixing effect. Moreover, variations in the flow velocity and blade angles are important factors affecting the system pressure drop. If the inlet velocity is 3 m/s and the blade angle is 90°, the static mixer exhibits optimized overall performance. The quantitative analysis shows that increasing the blade angle from 80° to 100° reduces the pressure drop by approximately 44%, while raising the inlet velocity from 3 m/s to 15 m/s lowers the outlet COV value by about 70%, indicating enhanced mixing uniformity. These findings confirm that an inlet velocity of 3 m/s combined with a 90° blade angle provides an optimal trade-off between mixing performance and energy efficiency. Full article

The viscoelasticity of fluids have a significant impact on the process of heat and mass transfer, which directly affects the efficiency and quality, especially for highly viscous functional polymer materials. In this work, the effect of elasticity on hydrodynamic behavior of pipe flow for highly viscous non-Newtonian fluids was studied using viscoelastic polyolefin elastomer (POE). Two constitutive rheological equations, the Cross model and Wagner model, were applied to describe the rheological behavior of typical POE melts, which have been embedded with computational fluid dynamics (CFD) simulation of the laminar pipe flow through the user-defined function (UDF) method. The influence of both viscosity and elasticity of a polymer melt on the flow mixing and heat transfer behavior has been systematically studied. The results show that the elastic effect makes a relative larger velocity gradient in the radial direction and the thicker boundary layer near pipe wall under the same feed flow rate. That leads to the higher pressure drop and more complex residence time distribution with the longer residence time near the wall but shorter residence time in the center. Under the same conditionals, the pipeline pressure drop of the viscoelastic fluid is several times or even tens of times greater than that of the viscous fluid. When the inlet velocity increases from 0.0001 m/s to 0.01 m/s, the difference in boundary layer thickness between the viscoelastic fluid and viscous fluid increases from 3% to 12%. Similarly, the radial temperature gradient of viscoelastic fluids is also relatively high. When the inlet velocity is 0.0001 m/s, the radial temperature difference of the viscoelastic fluid is about 40% higher than that of viscous fluid. Besides that, the influence of elasticity deteriorates the mixing effect of the SK type static mixer on the laminar pipe flow of highly viscous non-Newtonian fluids. Correspondingly, the accuracy of the simulation results was verified by comparing the pressure drop data from pipeline hydrodynamic experiments. Full article

With the accelerated transition of the global energy structure towards decarbonization, the share of PV power generation in the power system continues to rise. IEA predicts PV will account for 80% of new global renewable installations during 2025–2030. However, latent faults emerging from the long-term operation of photovoltaic (PV) power plants significantly compromise their operational efficiency. The existing EL detection methods in PV plants face challenges including grain boundary interference, probe band artifacts, non-uniform luminescence, and complex backgrounds, which elevate the risk of missing small defects. In this paper, we propose a high-precision defect detection method based on BiFDRep-YOLOv8n for small target defects in photovoltaic (PV) power plants, aiming to improve the detection accuracy and real-time performance and to provide an efficient solution for the intelligent detection of PV power plants. Firstly, the visual transformer RepViT is constructed as the backbone network, based on the dual-path mechanism of Token Mixer and Channel Mixer, to achieve local feature extraction and global information modeling, and combined with the structural reparameterization technique, to enhance the sensitivity of detecting small defects. Secondly, for the multi-scale characteristics of defects, the neck network is optimized by introducing a bidirectional weighted feature pyramid network (BiFPN), which adopts an adaptive weight allocation strategy to enhance feature fusion and improve the characterization of defects at different scales. Finally, the detection head part uses DyHead-DCNv3, which combines the triple attention mechanism of scale, space, and task awareness, and introduces deformable convolution (DCNv3) to improve the modeling capability and detection accuracy of irregular defects. Full article

Background/Objectives: Microfluidic mixing has become the gold standard procedure for manufacturing nucleic acid lipid-based delivery systems, offering precise control over critical process parameters. The choice and design of microfluidic mixers are often seen as a key driving force affecting the critical quality attributes of the resulting lipid nanoparticles (LNPs). Methods: This study aimed to evaluate LNPs manufactured using two low-cost microfluidic mixers alongside manual mixing (pipette mixing (PM)), followed by characterization studies using orthogonal analytics as well as expression studies to establish whether low-cost microfluidic manufacturing methods are suitable for bench-scale and high-throughput research. Results: The results show that all manufacturing methods can produce LNPs with sizes ranging between 95 and 215 nm with high encapsulation (70–100%), and enhanced analytics showed variations between the LNPs produced using the different mixers. Despite these differences, pipette mixing production of LNPs demonstrated its application as a high-throughput screening tool for LNPs, effectively distinguishing between different formulations and predicting consistent expression patterns both in vitro and in vivo. Conclusions: Overall, these results validate the use of low-cost microfluidic mixers without compromising the efficiency and integrity of the resulting LNPs. This study supports the increased accessibility of small-scale LNP manufacturing and high-throughput screening. Full article

The static mixer is one of the key equipment for dilution transportation of heavy oil. To enhance the mixing performance of heavy oil dilution, a static mixer featuring a parabolic blade has been developed through an innovative redesign of the traditional Kenics blade. Numerical simulations of the parabolic static mixer were conducted using Fluent 2022 R1 software. The coefficients of concentration variation (COV) and pressure drop (∆P) served as evaluation indexes, and the effects of parabolic focal length (P), torsion angle (α), and length–diameter ratio (Ar) of the mixing blade on mixing performance were thoroughly analyzed. The research indicates that setting the mixing blade parameters to P = 60, α = 180°, and Ar = 1.5 results in improved mixing performance compared to the traditional Kenics static mixer, achieving a COV of 0.036, which signifies nearly complete mixing of heavy oil and light oil. As parabolic P increases, ∆P exhibits a decreasing trend, while the COV begins to show a significant difference at the outlet of the third mixing blade. As α increases, ∆P rises, while the COV decreases. A decrease in Ar causes ∆P to increase sharply. Although heavy oil and light oil can mix rapidly over a short distance, their influence on the final mixing effect is relatively minor. This study offers significant theoretical insights and practical implications for high-efficiency heavy oil dilution transportation technology. Full article

The rapid advancement of terahertz (THz) communication systems has positioned this technology as a key enabler for next-generation telecommunication networks, including 6G, secure communications, and hybrid wireless-optical systems. This review comprehensively analyzes THz communication, emphasizing its integration with free-space optical (FSO) systems to overcome conventional bandwidth limitations. While THz-FSO technology promises ultra-high data rates, it is significantly affected by atmospheric absorption, particularly absorption beyond 500 GHz, where the attenuation exceeds 100 dB/km, which severely limits its transmission range. However, the presence of a lower-loss transmission window at 680 GHz provides an opportunity for optimized THz-FSO communication. This paper explores recent developments in high-power THz sources, such as quantum cascade lasers, photonic mixers, and free-electron lasers, which facilitate the attainment of ultra-high data rates. Additionally, adaptive optics, machine learning-based beam alignment, and low-loss materials are examined as potential solutions to mitigating signal degradation due to atmospheric absorption. The integration of THz-FSO systems with optical and radio frequency (RF) technologies is assessed within the framework of software-defined networking (SDN) and multi-band adaptive communication, enhancing their reliability and range. Furthermore, this review discusses emerging applications such as self-driving systems in 6G networks, ultra-low latency communication, holographic telepresence, and inter-satellite links. Future research directions include the use of artificial intelligence for network optimization, creating energy-efficient system designs, and quantum encryption to obtain secure THz communications. Despite the severe constraints imposed by atmospheric attenuation, the technology’s power efficiency, and the materials that are used, THz-FSO technology is promising for the field of ultra-fast and secure next-generation networks. Addressing these limitations through hybrid optical-THz architectures, AI-driven adaptation, and advanced waveguides will be critical for the full realization of THz-FSO communication in modern telecommunication infrastructures. Full article

Biogas, derived from human waste or industrial byproducts, is considered one of the most environmentally acceptable fuels. However, such fuels often exhibit relatively low efficiency, making it essential to develop technologies that facilitate their effective combustion. This article investigates the combustion of biogas with the addition of hydrogen at varying degrees of flow swirling. For this purpose, a burner was used in which methane, hydrogen and CO₂ were mixed in a mixer. The studies revealed that increasing the proportion of hydrogen in biogas leads to an average 15% rise in the NO_x concentration. Additionally, an increase in the degree of swirling has a positive effect on NO_x generation. On the other hand, a higher proportion of hydrogen reduces the concentration of CO in the exhaust gases. The presence of ballast gases, such as CO₂, generally results in relatively low NO_x levels when combined with a high swirling number. The analysis of combustion products for CO₂ indicates a 14% increase in CO₂ proportion. The highest concentrations of CO₂ were observed in biogas with the highest CO₂ ballast content. In terms of reducing NO_x and CO, SW = 1.3 is the most successful. On the other hand, this angle leads to an increase in the CO₂ concentration. Full article

Continuous mixers, which consist of a section with non-intermeshing counter-rotating rotors and a single-screw extruder, were developed for thermoset rubber and are often used for compounding of heavily filled thermoplastics. Due to the high mixing efficiency and tight control of shear levels, they may be suited for other compounding other material systems. Little work, however, has been reported on compounding with these mixers, and preliminary work with polypropylene showed interesting limitations of the mixing parameters. Therefore, this study investigated the effects of nine rotor designs, two single-screw designs, rotor speed, feed rate, and orifice setting on the residence time and melt temperature in a Farrel Compact Processor. In general, single-stage rotors produced lower mixer residence times and melt temperatures compared to longer two-stage and high dispersion rotors. Higher rotor speeds and feed rates and smaller orifice openings generally reduced mixer residence times. Higher rotor speeds increased mixer melt temperatures, whereas higher feed rates and smaller orifice openings produced lower mixer melt temperatures. The single-screw design impacted the residence time but not the melt temperature. Overall, the results of this work provided strategies for optimizing the processing parameters and rotor design selection when melt compounding with continuous mixers. Full article

Simple load forecasting and overload prediction models, such as LSTM and XGBoost, are unable to handle the increasing amount of data in power systems. Recently, various foundation models (FMs) for time series analysis have been proposed, which can be scaled up for large time series variables and datasets across domains. However, the simple pre-training setting makes FMs unsuitable for complex downstream tasks. Effectively handling real-world tasks depends on additional data, i.e., covariates, and prior knowledge. Incorporating these through structural modifications to FMs is not feasible, as it would disrupt the pre-trained weights. To address this issue, this paper proposes a frequency domain mixer, i.e., FreqMixer, framework for enhancing the task-specific analytical capabilities of FMs. FreqMixer is an auxiliary network for the backbone FMs that takes covariates as input. It has the same number of layers as the backbone and communicates with it at each layer, allowing the incorporation of prior knowledge without altering the backbone’s structure. Through experiments, FreqMixer demonstrates high efficiency and performance, reducing MAPE by 23.65%, recall by 87%, and precision by 72% in transformer load forecasting during the Spring Festival while improving precision by 192.09% and accuracy by 14% in corresponding overload prediction, all while processing data from over 160 transformers with just 1M additional parameters. Full article

The continuous synthesis of nanoparticles (NPs) has been actively studied due to its great potential to produce NPs with reproducible and controllable physicochemical properties. Here, we achieved the high throughput production of nanostructured lipid carriers (NLCs) using a coaxial turbulent jet mixer with an added heating system. This device, designed for the crossflow of precursor solution and non-solvent, combined with the heating system, efficiently dissolves solid lipids and surfactants. We reported the flow regime according to the Reynolds number (Re). Also, we confirmed the size controllability of NLCs as dependent on both Re and lipid concentration. The optimized synthesis yields NLCs around 80 nm, ideal for targeted drug delivery by enhanced permeability and retention (EPR) effect. The coaxial turbulent jet mixer enables effective mixing, producing uniform size distribution of NLCs. The NLCs prepared using the coaxial turbulent jet mixer were smaller, more uniform, and had higher drug loading compared to the NLCs synthesized by a bulk nanoprecipitation method, showcasing its potential for advancing nanomedicine. Full article