Experimental Evaluation Research Articles

A novel personalized homogenous finite element model to predict the pull-out strength of cancellous bone screws.

Orthopedic surgeries often involve the insertion of bone screws for various fixation systems. The risk of postoperative screw loosening is usually assessed through experimental or finite element (FE) evaluations of the screw pull-out strength. FE simulations are based on either personalized complex but accurate heterogeneous modeling or non-personalized simple but relatively less accurate homogeneous modeling. This study aimed to develop and validate a novel personalized computed tomography (CT)-based homogeneous FE simulation approach to predict the pull-out force of cancellous bone screws. Twenty FE simulations of L1-L5 vertebral screw pull-out tests were conducted, i.e., 10 heterogeneous and 10 homogenous models. Screws were inserted into the lower-middle region of vertebrae. In our novel homogeneous model, the region around approximately twice the diameter of the screw was used as a bone material reference volume. Subsequently, the overall material property of this region was homogeneously attributed to the entire vertebra, and pull-out simulations were conducted. The mean error of the predicted pull-out forces by our novel homogenous simulations was ~ 7.9% with respect to our heterogeneous model. When solely the cancellous bone was involved during the pull-out process (i.e., for L1, L2, and L3 vertebral bodies whose cortical bone in the inferior region is thin), the novel homogenous model yielded small mean error of < 6.0%. This error, however, increased to ~ 11% when the screw got involved to the cortical bone (for L4 and L5 vertebrae whose cortical bone in the inferior region is thick). The proposed personalized CT-based homogenous model was highly accurate in estimating the pull-out force especially when only the cancellous bone was involved with the screw.

Visual software defect prediction method based on improved recurrent criss-cross residual network

Purpose This study aims to solve the problems of large training sample size, low data sample quality, low efficiency of the currently used classical model, high computational complexity of the existing concern mechanism, and high graphics processing unit (GPU) occupancy in the current visualization software defect prediction, proposing a method for software defect prediction termed recurrent criss-cross attention for weighted activation functions of recurrent SE-ResNet (RCCA-WRSR). First, following code visualization, the activation functions of the SE-ResNet model are replaced with a weighted combination of Relu and Elu to enhance model convergence. Additionally, an SE module is added before it to filter feature information, eliminating low-weight features to generate an improved residual network model, WRSR. To focus more on contextual information and establish connections between a pixel and those not in the same cross-path, the visualized red as integer, green as integer, blue as integer images are inputted into a model incorporating a fused RCCA module for defect prediction. Design/methodology/approach Software defect prediction based on code visualization is a new software defect prediction technology, which mainly realizes the defect prediction of code by visualizing code as image, and then applying attention mechanism to extract the features of image. However, the challenges of current visualization software defect prediction mainly include the large training sample size and low sample quality of the data, and the classical models used today are not efficient, and the existing attention mechanisms have high computational complexity and high GPU occupancy. Findings Experimental evaluation using ten open-source Java data sets from PROMISE and five existing methods demonstrates that the proposed approach achieves an F-measure value of 0.637 in predicting 16 cross-version projects, representing a 6.1% improvement. Originality/value RCCA-WRSR is a new visual software defect prediction based on recurrent criss-cross attention and improved residual network. This method effectively enhances the performance of software defect prediction.

Few‐Shot Contrastive Learning‐Based Multi‐Round Dialogue Intent Classification Method

ABSTRACTTraditional text classification models face challenges in handling long texts and understanding topic transitions in dialogue scenarios, leading to suboptimal performance in automatic speech recognition (ASR)‐based multi‐round dialogue intent classification. In this article, we propose a few‐shot contrastive learning‐based multi‐round dialogue intent classification method. First, the ASR texts are partitioned, and role‐based features are extracted using a Transformer encoder. Second, refined sample pairs are forward‐propagated, adversarial samples are generated by perturbing word embedding matrices and contrastive loss is applied to positive sample pairs. Then, positive sample pairs are input into a multi‐round reasoning module to learn semantic clues from the entire scenario through multiple dialogues, obtain reasoning features, input them into a classifier to obtain classification results, and calculate multi‐task loss. Finally, a prototype update module (PUM) is introduced to rectify the biased prototypes by using gated recurrent unit (GRU) to update the prototypes stored in the memory bank and few‐shot learning (FSL) task. Experimental evaluations demonstrate that the proposed method outperforms state‐of‐the‐art methods on two public datasets (DailyDialog and CM) and a private real‐world dataset.

Characteristics of a novel elliptical ultrasonic vibration-assisted turning device

ABSTRACT 2D Ultrasonic Vibration-Assisted Turning (2D-UVAT) is a proven method to improve the traditional turning process. This process utilises bi-axial vibrations to enhance the smoothness of the machined surface and to suppress cutting temperature and cutting force. This study investigates the effects of employing a novel non-resonant vibration module, called the RNO vibrator, on the surface roughness, cutting temperature, and cutting force of aluminium alloy 6061-T6 machined using 2D-UVAT technique. The RNO vibrator operates by harnessing flexural hinges to disseminate and amplify the actuated vibrations originating from two orthogonal piezo actuator stacks. This vibrator was designed from a simple bar concept, expecting minimal adjustment on the tool post during installation. The experimental evaluation of the RNO vibrator’s performance was conducted by varying key machining parameters, including spindle speed, depth of cut, feed rate, and vibrational frequency. The study reveals that the introduced RNO vibrator substantially enhances machining quality by diminishing surface roughness and cutting temperature, while concurrently reducing cutting forces. Under the given conditions, the machined workpieces had a surface roughness between 0.41 and 1.25 µm and a cutting temperature ranging from 153.36°C to 180.5°C, while the cutting force was low, typically between 35.3 N and 145.5 N. The vibrator works better at high frequency, particularly at 20 kHz. Eventually, the study suggests that the RNO vibrator has a promising performance as a vibration module for the 2D-UVAT process.

On the mechanism for work function change of gold electrodes by ultrathin polyethyleneimine (PEI) films: Effect of molecular order.

Polyethyleneimine (PEI) is a widely used cationic polyelectrolyte. In organic electronics, it is a universal surface modifier for shifting the electrode work function (Φ) and improving charge injection into electronic devices. This effect may depend on the conformation and dipolar order of the PEI ultrathin film, but their detailed experimental evaluation has not yet been reported. Thus, we used sum-frequency generation (SFG) spectroscopy to probe the net orientation of polar groups of PEI films on glass and gold. The films were fabricated by spin-coating from alcoholic solutions or by dip-coating from aqueous solutions of various pH values, with both branched (b-PEI) and linear (l-PEI) structures. The obtained SFG spectra and atomic force microscopy (AFM) images indicated that the conformational ordering of the PEI layers increases over the period of 14 days after fabrication, being slightly more pronounced for l-PEI vs b-PEI, and for dip-coating vs spin coating fabrication. Furthermore, both the pH of the dip-coating solutions and the substrate nature influence the final morphology and order of the adsorbed films. On glass, they are optimized at an intermediate pH 5, while on gold, the greatest homogeneity is observed at pH 2 and the largest dipolar order is observed at pH 10. The pH dependence of changes in the work function of gold by PEI (|ΔΦ|) suggests that the electronic contribution is dominant. Nevertheless, the evolution of the PEI dipolar ordering was accompanied by small variations of |ΔΦ|, suggesting that it does have a significant contribution, especially at conditions for which the electronic contribution is reduced.

Recognizing human activities using light-weight and effective machine learning methodologies

Background Human activity recognition (HAR) is increasingly important in enhancing healthcare systems by enabling accurate monitoring of individuals' movements through sensor data. This paper is motivated by the need to improve the accuracy of HAR, particularly for applications in e-health systems, where reliable activity detection can lead to better health outcomes. The study explores six prominent machine learning techniques—decision tree, random forest, linear regression, Naïve Bayes, k-nearest neighbor, and neural networks—to determine which methods can most effectively predict activities like walking, sitting, standing, laying, walking upstairs, and walking downstairs. Methods We employed these six machine learning algorithms to analyze a comprehensive dataset derived from various sensors. Each model was rigorously trained and evaluated to compare its effectiveness in recognizing human activities. The experiments aimed to identify strengths and weaknesses in each approach, with particular emphasis on advanced techniques such as random forest, convolutional neural networks (CNNs), and gated recurrent networks (GRNs). Results The experimental evaluation revealed that the random forest classifier, CNN, GRN, and neural networks delivered promising results, achieving high accuracy levels. Notably, the neural network model excelled, attaining an impressive accuracy of 98%. In contrast, the Naïve Bayes model did not meet the performance expectations set by the other algorithms. Conclusions This research effectively classifies activities such as sitting, standing, laying, walking, walking downstairs, and walking upstairs, underscoring the potential of machine learning in HAR. The findings highlight the superior performance of neural networks in enhancing activity recognition, which could lead to advanced applications in e-health systems and improve overall healthcare monitoring strategies.

A novel light GBM framework combined with gain optimization for predicting dementia severity of Alzheimer’s disease

ABSTRACT Introduction Alzheimer’s disease (AD) is a severe brain disorder, which comprises almost 75% of all types of dementia cases. As per the latest report, the global count of Alzheimer’s disease patients is 131 million, due to which the treatment of Alzheimer’s disease is becoming a major burden to the world healthcare system. However, Alzheimer’s disease still remains incurable due to its multifactorial nature of symptoms. Therefore, early-stage diagnosis of Alzheimer’s disease is essential, which helps in the treatment and recovery of patients to a greater extent. Background In the existing literature, few research attempts are carried out for Alzheimer’s disease prediction using machine learning algorithms; yet, most of the existing techniques fail to focus on aspects such as gain-optimisation, dementia severity identification and accuracy enhancement. Methodology To solve these issues, this article presents a light gradient boosting machine-based framework, which employs a gain-optimisation algorithm using multi-category factors for the early stage prediction of Alzheimer’s disease. Further, the proposed framework introduces a detailed feature analysis for identifying the key features needed for the disease prediction and thereby classifies the Alzheimer’s disease patients into different dementia severity types. Results The extensive evaluations of the proposed framework on different dataset sizes using real-world MRI datasets of patients, clearly demonstrate the performance of the proposed work in terms of gain-optimised accuracy metrics. Further, experimental evaluations clearly prove the better performance of the proposed framework when compared to that of other ML techniques including LR, RFC, SVR, SVC and existing methods in terms of Recall, F1-score and accuracy metrics.

Experimental Evaluation of Under Slab Mats (USMs) Made from End-of-Life Tires for Ballastless Tram Track Applications

The growing population of urban areas results in the need to deal with the noise pollution from the transportation system. This study presents experimental test results of static and dynamic elastic characteristics of under slab mats (USMs) according to the procedure of DIN 45673-7. Prototype USMs based on recycled elastomeric materials, i.e., SBR granules and fibres produced from waste car tires, are analysed. Vibration isolation mats with different thicknesses (10, 15, 20, 25, 30, and 40 mm), densities (500 and 600 kg/m3), and different degrees of space filling (no holes, medium holes, large holes) are considered. Moreover, a practical application of the laboratory test results of USMs in the design of ballastless track structures of two different types (with a concrete slab and longitudinal beams) is presented. Deflections of the rail and the floating slab system, as well as stresses acting on the mat, are determined according to EN 16432-2. The use of shredded rubber from recycled car tires as a material component of sustainable and environmentally friendly tram track structures may be one of the most effective ways to manage rubber waste within the current trend toward a circular economy, and this study intends to introduce methods for experimental identification and analytical selection of basic static and dynamic parameters of prototype USMs.

Experimental Evaluation of Gas-Dynamic Conditions of Heat Exchange of Stationary Air Flows in Vertical Conical Diffuser

Conical diffusers are widely used in technical devices (gasifiers, turbines, combustion chambers) and technological processes (ejectors, mixers, renewable energy). The perfection of flow gas dynamics in a conical diffuser affects the intensity of heat and mass transfer processes, the quality of mixing/separation of working media and the flow characteristics of technical devices. These parameters largely determine the efficiency and productivity of the final product. This article presents an analysis of experimental data on the gas-dynamic characteristics of stationary air flows in a vertical, conical, flat diffuser under different initial boundary conditions. An experimental setup was created, measuring instruments were selected, and an automated data collection system was developed. Basic data on the gas dynamics of air flows were obtained using the thermal anemometry method. Experimental data on instantaneous values of air flow velocity in a diffuser for initial velocities from 0.4 m/s to 2.22 m/s are presented. These data were the basis for calculating and obtaining velocity fields and turbulence intensity fields of the air flow in a vertical diffuser. It is shown that the value of the initial flow velocity at the diffuser inlet has a significant effect on the gas-dynamic characteristics. In addition, a spectral analysis of the change in air flow velocity both by height and along the diffuser axis was performed. The obtained data may be useful for refining engineering calculations, verifying mathematical models, searching for technical solutions and deepening knowledge about the features of gas dynamics of air flows in vertical diffusers.

Research on Personalized Cultural Learning Platform Based on Collaborative Filtering and Popularity Recommendation

With the rapid progress of information technology, it has become the demand of the times to use the power of the Internet to inherit and develop Chinese traditional culture. Given the challenge of users searching for learning content in massive online resources, personalized recommendation systems have become the key to improving user experience. This paper focuses on building a personalized Chinese traditional culture learning platform based on collaborative filtering and popularity recommendation, aiming to stimulate users’ interest in learning traditional culture through intelligent algorithms. In terms of technical implementation, this study relies on the Django framework and MySQL database to build a feature-rich and user-friendly traditional cultural learning platform, supporting users’ diverse learning needs, and providing flexible learning paths and abundant learning resources. Experimental evaluation shows that the hybrid algorithm of collaborative filtering and popularity recommendation proposed in this paper has improved recommendation performance compared to existing methods, effectively enhancing user satisfaction and platform activity. This platform not only promotes the digital dissemination of traditional culture, but also provides users with a highly autonomous and personalized learning experience, meeting the contemporary society’s pursuit of diversified and personalized cultural learning.

Effect of Delaying High School Start Time on Teen Physical Activity, Screen Use, and Sports and Extracurricular Activity Participation: Results From START.

We aimed to characterize relationships between delayed high school start time policy, which is known to lengthen school night sleep duration, and patterns in activity outcomes: physical activity, non-school electronic screen time (non-schoolwork), and sports and extracurricular activity among adolescents. We used data from the START study, a multi-site evaluation of a natural experiment, assessing the effects of a school start time policy change in high schools in the Minneapolis, Minnesota metropolitan area. The study follows students in 2 schools that shifted to a later start time (8:20 or 8:50 am) after baseline year and 3 schools that maintained a consistent, early start time (7:30 am) over the 3-year study period. Activity was measured by participant self-report on an in-school survey. The analysis used a difference-in-differences estimator, in which the changes in each outcome observed in the comparison schools estimate the changes in each outcome that would have been observed in the late-start adopting schools had they not delayed their start times after baseline. Over 2 years of follow-up, no changes emerged to suggest that later school start times either interfered with, or promoted, any activity-related outcome that was measured. Communities interested in promoting sleep by delaying start times may do so knowing that there are unlikely to be adverse effects on adolescent physical activity, electronic screen time, or organized sports and activity participation. A shift to later school start times does not appear to enhance or detract from the healthfulness of students' activity level.

Adaptation of Object Detection Algorithms for Road Indicator Lights in Complex Scenes

In the realm of autonomous driving, practical driving scenarios are fraught with numerous complexities, including inclement weather conditions, nighttime blurriness, and ambient light sources that significantly hinder drivers’ ability to discern road indicators. Furthermore, the dynamic nature of road indicators, which are constantly evolving, poses additional challenges for computer vision-based detection systems. To address these issues, this paper introduces a road indicator light detection model, leveraging the advanced capabilities of YOLOv8. We have ingeniously integrated the robust backbone of YOLOv8 with four distinct attention mechanism modules—Convolutional Block Attention Module (CBAM), Efficient Channel Attention (ECA), Shuffle Attention (SA), and Global Attention Mechanism (GAM)—to significantly enhance the model performance in capturing nuanced features of road indicators and boosting the accuracy of detecting minute objects. Additionally, we employ the Asymptotic Feature Pyramid Network (AFPN) strategy, which optimizes the fusion of features across different scales, ensuring not only an enhanced performance but also maintaining real-time capability. These innovative attention modules empower the model by recalibrating the significance of both channel and spatial dimensions within the feature maps, enabling it to hone in on the most pertinent object characteristics. To tackle the challenges posed by samples rich in small, occluded, background-similar objects, and those that are inherently difficult to recognize, we have incorporated the Focaler-IOU loss function. This loss function deftly reduces the contribution of easily detectable samples to the overall loss, thereby intensifying the focus on challenging samples. This strategic balancing of hard-to-detect versus easy-to-detect samples effectively elevates the model’s detection performance. Experimental evaluations conducted on both a public traffic signal dataset and a proprietary headlight dataset have yielded impressive results, with both mAP50 and mAP50:95 metrics experiencing significant improvements exceeding two percentage points. Notably, the enhancements observed in the headlight dataset are particularly profound, signifying a significant step forward toward realizing safer and more reliable assisted driving technologies.

Multi-Epitopic Peptide Vaccine Against Newcastle Disease Virus: Molecular Dynamics Simulation and Experimental Validation

Background: Newcastle disease virus (NDV) is a highly contagious and economically devastating pathogen affecting poultry worldwide, leading to significant losses in the poultry industry. Despite existing vaccines, outbreaks continue to occur, highlighting the need for more effective vaccination strategies. Developing a multi-epitopic peptide vaccine offers a promising approach to enhance protection against NDV. Objectives: Here, we aimed to design and evaluate a multi-epitopic vaccine against NDV using molecular dynamics (MD) simulation. Methodology: We retrieved NDV sequences for the fusion (F) protein and hemagglutinin–neuraminidase (HN) protein. Subsequently, B-cell and T-cell epitopes were predicted. The top potential epitopes were utilized to design the vaccine construct, which was subsequently docked against chicken TLR4 and MHC1 receptors to assess the immunological response. The resulting docked complex underwent a 1 microsecond (1000 ns) MD simulation. For experimental evaluation, the vaccine’s efficacy was assessed in mice and chickens using a controlled study design, where animals were randomly divided into groups receiving either a local ND vaccine or the peptide vaccine or a control treatment. Results: The 40 amino acid peptide vaccine demonstrated strong binding affinity and stability within the TLR4 and MHC1 receptor–peptide complexes. The root mean square deviation of peptide vaccine and TLR4 receptor showed rapid stabilization after an initial repositioning. The root mean square fluctuation revealed relatively low fluctuations (below 3 Å) for the TLR4 receptor, while the peptide exhibited higher fluctuations. The overall binding energy of the peptide vaccine with TLR4 and MHC1 receptors amounted to −15.7 kcal·mol−1 and −36.8 kcal·mol−1, respectively. For experimental evaluations in mice and chicken, the peptide vaccine was synthesized using services of GeneScript Biotech® (Singapore) PTE Limited. Experimental evaluations showed a significant immune response in both mice and chickens, with the vaccine eliciting robust antibody production, as evidenced by increasing HI titers over time. Statistical analysis was performed using an independent t-test with Type-II error to compare the groups, calculating the p-values to determine the significance of the immune response between different groups. Conclusions: Multi-epitopic peptide vaccine has demonstrated a good immunological response in natural hosts.

Text-video retrieval re-ranking via multi-grained cross attention and frozen image encoders

State-of-the-art methods for text-video retrieval generally leverage CLIP embeddings and cosine similarity for efficient retrieval. Meanwhile, recent advancements in cross-attention techniques introduce transformer decoders to facilitate attention computation between text queries and visual tokens extracted from video frames, enabling a more comprehensive interaction between textual and visual information. In this study, we combine the advantages of both approaches and propose a fine-grained re-ranking approach incorporating a multi-grained text-video cross attention module. Specifically, the re-ranker enhances the top K similar candidates identified by the cosine similarity network. To explore video and text interactions efficiently, we introduce frame and video token selectors to obtain salient visual tokens at both frame and video levels. Then, a multi-grained cross-attention mechanism is applied between text and visual tokens at these levels to capture multimodal information. To reduce the training overhead associated with the multi-grained cross-attention module, we freeze the vision backbone and only train the multi-grained cross attention module. This frozen strategy allows for scalability to larger pre-trained vision models such as ViT-G, leading to enhanced retrieval performance. Experimental evaluations on text-video retrieval datasets showcase the effectiveness and scalability of our proposed re-ranker combined with existing state-of-the-art methodologies.

Wearing myopia glasses on GANs: Mitigating bias for pre-trained Generative Adversarial Networks via online prior perturbation

Pre-trained Generative Adversarial Networks (GANs) can provide rich information and make various downstream tasks beneficial. However, the training process of GANs potentially learns pre-existing biases in the historical data, which are not only inherited, but also amplified to some extent when using the generated data for prediction. Most available methods for bias mitigation utilize pre-processing or in-processing strategies to retrain GANs, which needs to access the raw data and also increases the time and computational cost required to train a fully expressive GAN. Therefore, we propose a novel post-processing mechanism to achieve fairness outputs of GANs through online prior perturbation. Unlike traditional output-based offline processing, our method mitigates biases from model inputs by designing a prior perturbation network, called “prior perturber”, to form a combined network with the pre-trained GAN. Specifically, we introduce a bias prediction network, called “bias predictor”, for online adversarial training with the prior perturber, which can decouple the generated representations from the bias features and improve the fairness of the downstream prediction results. In addition, the training process is independent of specific target labels, and the generated fair representations have good transferability. Experimental evaluations on five real datasets validate the effectiveness of the proposed method and the best utility is obtained on the same fairness level as compared to other output-based baseline methods. The result on fairness-utility trade-offs significantly improves the fairness of original samples by 56.40%, while the utility is reduced by only 2.60%.

Solanum tuberosum leaf extract as an eco-friendly corrosion inhibitor for Q235-steel in HCl medium: experimental and theoretical evaluation

Solanum tuberosum leaf extract (STLs), used as a biodegradable corrosion inhibitor, was prepared to mitigate the corrosion rate of Q235 steel in HCl medium. The active inhibitory components existing in STLs were tracked by Fourier transform infrared spectroscopy. Electrochemical measurements, morphological characterization and contact angle tests were combined to evaluate the anti-corrosion performance of STLs at various concentrations and different temperatures, and the results show that the addition of STLs can evidently magnify the charge transfer resistance, expand the double layer capacitance at electric /solution interface, and simultaneously reduce the corrosion current density of Q235 steel in HCl medium with the maximum inhibition efficiency of 91.89%. The adsorption behavior of STLs on steel was investigated using X-ray photoelectron spectroscopy and adsorption models. It is revealed that the molecules of STLs spontaneously bind to the substrate through chemical and physical action, forming a protective layer to block the diffusion pathway of harsh ions. Furthermore, theoretical calculations confirm that Solanum tuberosum leaves components were adsorbed on steel substrates through donor-acceptor interactions.

Experimental and model evaluation of cyclic compressive performance for RAC-filled BFRP tube composite columns

The mechanical properties of recycled aggregate concrete (RAC) can be significantly enhanced in RAC-filled fiber-reinforced-polymer (FRP) tube composite columns. However, the cyclic compressive behavior of this type of composite columns has been scarcely studied. In this study, a series of cyclic axial compressive tests were conducted on RAC-filled basalt-FRP (BFRP) tube composite columns, and the effects of confinement level (number of BFRP layers) and RA replacement ratio were investigated. The test results indicate that, compared to RAC columns, the compressive capacity and deformability of RAC-filled BFRP tube composite columns increased significantly under cyclic loading. The compressive strength and ultimate strain of BFRP-confined RAC reached 1.42-3.98 times and 2.88-21.53 times those of unconfined RAC, respectively. The compressive strength and ultimate strain under cyclic loading were generally close to those under monotonic loading, with the compressive strength ratios averaging 1.052 and the ultimate strain ratios averaging 1.031. The enhancement efficiency of BFRP confinement increased progressively with increasing confinement level and/or RA replacement ratio. The envelope curves of the cyclic stress-strain curves resembled the stress-strain curves under monotonic loading. The reloading modulus (Ere) and residual modulus (Eun,0) continuously decreased as the unloading strain (εun) increased, while the plastic strain (εpl) increased almost linearly with εun. The values of Ere, Eun,0 and εpl were affected by the confinement level and RA replacement ratio. Finally, typical theoretical models for FRP-confined concrete were used to evaluate the cyclic stress-strain curves of RAC-filled BFRP tube composite columns.

Comprehensive evaluation of solar floating photovoltaic prospective in Saudi Arabia: Comparative experimental investigation and thermal performance analysis

Recently, floating photovoltaic systems have been regarded as a promising technology for producing clean energy by utilizing the surface of water bodies, such as lakes, rivers and oceans. This study introduces a comparative experimental study and energy performance evaluation of a 1.0 kW offshore floating photovoltaic (FPV) system and a nearby traditional ground-based PV system (GPV) installed in the eastern province of Saudi Arabia. The FPV system was deployed in the Arabian Gulf, 25 m off the coast, at an average depth of 1 to 1.5 m depending on tide, wave, and current intensity. The FPV system employs a strong novel eco-friendly platform structure made of recycled buoyant materials such as engineered plastic drums and wood. This system is anchored with tension cables and concrete blocks that can withstand the changing sea water conditions. The GPV system, on the other hand, was installed 150 m inland from the shore. Real-time monthly data monitoring and assessment indicated that FPV systems outperformed GPV systems in terms of lower PV panel surface temperatures, higher power output, and panel efficiency. Based on daily average back surface temperatures, FPV system temperature was decreased by 7.5 % to 21.34 % when compared to GPV. Moreover, the FPV system efficiency was also increased by 12.2 % when compared to the GPV system. This study aims to assist in promoting the applicability of solar floating photovoltaic systems to synergistically fulfill the requirements of sustainable electricity production for the arid costal community in Saudi Arabia and similar areas.

Multi‐omics graph convolutional networks for digestive system tumour classification and early‐late stage diagnosis

AbstractThe prevalence of digestive system tumours (DST) poses a significant challenge in the global crusade against cancer. These neoplasms constitute 20% of all documented cancer diagnoses and contribute to 22.5% of cancer‐related fatalities. The accurate diagnosis of DST is paramount for vigilant patient monitoring and the judicious selection of optimal treatments. Addressing this challenge, the authors introduce a novel methodology, denominated as the Multi‐omics Graph Transformer Convolutional Network (MGTCN). This innovative approach aims to discern various DST tumour types and proficiently discern between early‐late stage tumours, ensuring a high degree of accuracy. The MGTCN model incorporates the Graph Transformer Layer framework to meticulously transform the multi‐omics adjacency matrix, thereby illuminating potential associations among diverse samples. A rigorous experimental evaluation was undertaken on the DST dataset from The Cancer Genome Atlas to scrutinise the efficacy of the MGTCN model. The outcomes unequivocally underscore the efficiency and precision of MGTCN in diagnosing diverse DST tumour types and successfully discriminating between early‐late stage DST cases. The source code for this groundbreaking study is readily accessible for download at https://github.com/bigone1/MGTCN.

Experimental Evaluation Study of the Comfort and Energy Performance of Suspended Particle Device Smart Windows in a Residential Building: Achieving Optimal Control during the Cooling Season

In contrast to traditional windows, smart windows dynamically affect indoor visual and thermal environments through Visible light transmittance (VT) and Solar heat gain coefficient (SHGC) changes. Therefore, for enhancing smart window application effects, it is necessary to analyze comprehensively the impacts of smart window tinting on indoor visual and thermal environments, and to establish appropriate tinting control methods. This study aimed to determine optimal smart window control strategies for residential buildings during the cooling season by evaluating visual and thermal comfort and energy performance in an actual building equipped with SPD smart windows. An experimental building was located in Sejong, South Korea (36.5°N latitude, 127.3°E longitude; temperate climate zone), and the measurements were conducted from July to August 2023 to compare glare, indoor daylight illuminance, thermal comfort, and cooling and lighting energy between typical and smart windows. And, optimal control methods were proposed from general comfort and energy performance perspectives. The measurements indicated that it was difficult to satisfy all aspects of the visual and thermal environment, and comprehensively it was recommended to employing “always tinting” on summer. However, on a clear day, jointly applying smart windows and lighting dimming control and controlling automatically based on indoor daylight illuminance levels were found to be the most effective as it not only secured visual comfort performance but also reduced lighting energy consumption.