Eng

With the large-scale closure of coal mines leading to groundwater pollution, in order to systematically identify the sources of major chemical ions in surface water and groundwater. This study comprehensively applied methods such as Piper’s trilinear diagram, linear fitting, and correlation analysis to quantitatively analyze the hydrochemical characteristics of closed coal mining areas in southwest Shandong and to clarify the sources of geochemical components in surface water and groundwater, and the PMF model was used to analyze the sources of chemical components in mine water and karst water. The results show that the concentrations of TDS ( Total Dissolved Solids), SO₄²⁻, Fe, and Mn in the mine water of the closed coal mine area are higher than in the karst water. Both water bodies are above groundwater quality standards. Ca²⁺, SO₄²⁻, and HCO₃⁻ dominate the ionic components in surface water and different types of groundwater. The hydrochemical types of surface, pore, and mine waters are mainly SO₄-HCO₃-Ca, whereas SO₄-HCO₃-Ca and HCO₃-SO₄-Ca dominate karst waters. SO₄²⁻ is the leading ion in the TDS of water bodies. The mineralization process of surface water is mainly controlled by the weathering of silicate minerals, while that of the groundwater is mainly controlled by the dissolution of carbonate minerals. The impact of mining activities on surface water and groundwater is significant, while the impact of agricultural activities on surface water and groundwater is relatively small. The degree of impact of coal mining activities on SO₄²⁻ concentrations in surface water, pore water, and karst water, in descending order, is karst water, surface water, and pore water. The PMF (Positive Matrix Factorization) model analysis results indicate that dissolution of carbonate minerals with sulphate and oxidation dissolution of sulfide minerals are the main sources of chemical constituents in mine waters. Carbonate dissolution, oxidation dissolution of sulfide minerals, domestic sewage, and dissolution of carbonate minerals with sulphate are ranked as the main sources of chemical constituents in karst water from highest to lowest. These findings provide a scientific basis for the assessment and control of groundwater pollution in the areas of closed coal mines. Full article

Dementia is a neurodegenerative disorder characterized by progressive cognitive impairment that significantly affects daily living. Early detection of Alzheimer’s disease—the most common form of dementia—remains essential for prompt intervention and treatment, yet clinical diagnosis often requires extensive and resource-intensive procedures. This article explores the effectiveness of automated Natural Language Processing (NLP) methods for identifying Alzheimer’s indicators from audio transcriptions of the Cookie Theft picture description task in the PittCorpus dementia database. Five NLP approaches were compared: a classical Tf–Idf statistical representation and embeddings derived from large language models (GloVe, BERT, Gemma-2B, and Linq-Embed-Mistral), each integrated with a logistic regression classifier. Transcriptions were carefully preprocessed to preserve linguistically relevant features such as repetitions, self-corrections, and pauses. To compare the performance of the five approaches, a stratified 5-fold cross-validation was conducted; the best results were obtained with BERT embeddings (84.73% accuracy) closely followed by the simpler Tf–Idf approach (83.73% accuracy) and the state-of-the-art model Linq-Embed-Mistral (83.54% accuracy), while Gemma-2B and GloVe embeddings yielded slightly lower performances (80.91% and 78.11% accuracy, respectively). Contrary to initial expectations—that richer semantic and contextual embeddings would substantially outperform simpler frequency-based methods—the competitive accuracy of Tf–Idf suggests that the choice and frequency of the words used might be more important than semantic or contextual information in Alzheimer’s detection. This work represents an effort toward implementing user-friendly software capable of offering an initial indicator of Alzheimer’s risk, potentially reducing the need for an in-person clinical visit. Full article

To analyze the wear performance induced by rotor–stator rubbing in an aero-engine sealing structure under authentic operating conditions, a transonic rotor system with double bearing is constructed. This system incorporates the disk, shaft, blades, joint bolts, and auxiliary support structure. The system was evaluated in terms of its critical speed, vibration characteristics, component strength under operational conditions, and response characteristics in abnormal extreme scenarios. A ball screw-type feeding system is employed to achieve precise rotor–stator rubbing during rotation by controlling the coating feed. Additionally, a quartz lamp heating system is used to apply thermal loads to coating specimens, and the appropriate heat insulation and cooling measures are implemented. Furthermore, a high-frequency rubbing force test platform is developed to capture the key characteristics caused by rubbing. The test rig can conduct response tests of the system with rotor–stator rubbing and abrasion tests with tip speeds reaching 425 m/s, feed rates ranging from 2 to 2000 μm/s, and heating temperatures up to 1200 °C. Test debugging has confirmed these specifications and successfully executed rubbing tests, which demonstrate stability throughout the process and provide reliable rubbing force test results. This designed test rig and analysis methodology offers valuable insights for developing high-speed rotating machinery. Full article

In today’s fast-paced world of software development, it is essential to ensure that programs run smoothly without any issues. When dealing with complex applications, the objective is to predict and resolve problems before they escalate. The prediction of software defects is a crucial element in maintaining the stability and reliability of software systems. This research addresses this need by combining advanced techniques (ensemble techniques) with seventeen machine learning algorithms for predicting software defects, categorised into three types: semi-supervised, self-supervised, and supervised. In supervised learning, we mainly experimented with several algorithms, including random forest, k-nearest neighbors, support vector machines, logistic regression, gradient boosting, AdaBoost classifier, quadratic discriminant analysis, Gaussian training, decision tree, passive aggressive, and ridge classifier. In semi-supervised learning, we tested are autoencoders, semi-supervised support vector machines, and generative adversarial networks. For self-supervised learning, we utilized are autoencoder, simple framework for contrastive learning of representations, and bootstrap your own latent. After comparing the performance of each machine learning algorithm, we identified the most effective one. Among these, the gradient boosting AdaBoost classifier demonstrated superior performance based on an accuracy of 90%, closely followed by the AdaBoost classifier at 89%. Finally, we applied ensemble methods to predict software defects, leveraging the collective strengths of these diverse approaches. This enables software developers to significantly enhance defect prediction accuracy, thereby improving overall system robustness and reliability. Full article

This study aimed to develop and characterize biodegradable films based on potato starch reinforced with carboxymethylcellulose (CMC) and polyethylene oxide (PEO) nanofibers, with the goal of improving their mechanical and thermal properties for potential use in sustainable packaging. The films were prepared through the thermal gelatinization of starch extracted from tubers, combined with nanofibers obtained by electrospinning CMC synthesized from potato starch. Key electrospinning variables, including solution concentration, voltage, distance, and flow rate, were analyzed. The films were morphologically characterized using scanning electron microscopy (SEM) and chemically analyzed by Fourier Transform Infrared Spectroscopy (FTIR) and X-ray Diffraction (XRD), and their thermal properties were assessed by Thermogravimetric Analysis (TGA) and Differential Scanning Calorimetry (DSC). The results indicated an increase in tensile strength to 14.1 MPa in the reinforced films, compared to 13.6 MPa for pure starch and 7.1 MPa for the fiber-free CMC blend. The nanofibers had an average diameter of 63.3 nm and a porosity of 32.78%. A reduction in crystallinity and more stable thermal behavior were also observed in the composite materials. These findings highlight the potential of using agricultural waste as a functional reinforcement in biopolymers, providing a viable and environmentally friendly alternative to synthetic polymers. Full article

Geopolymers (GPs), as promising alternatives to ordinary Portland cement (OPC)-based concrete, have gained interest in the last 20 years due to their enhanced mechanical properties, durability, and lower environmental impact. Synthesized from industrial by-products such as slag and fly ash, geopolymers offer a sustainable solution to waste management, resource utilization, and carbon dioxide reduction. However, similarly to OPC, geopolymers exhibit brittle behavior, and this characteristic defines a limit for structural applications. To tackle this issue, researchers have focused on the characterization, development, and implementation of fiber-reinforced geopolymers (FRGs), which incorporate various fibers to enhance toughness, ductility, and crack resistance, allowing their use in a wide range of structural applications. Following a general overview of sustainability considerations, this review critically analyzes the structural performance and capability of geopolymers in structural repair applications. Geopolymers demonstrate notable potential in new construction and repair applications. However, challenges such as complex mix designs, the availability of alkaline activators, curing temperatures, fiber matrix compatibility issues, and limited standards are restricting its large-scale adoption. The analysis and consolidation of an extensive dataset would support the viability of geopolymer as a durable and sustainable alternative to what is currently used in the construction industry, especially when fiber reinforcement is effectively integrated. Full article

Mango, as an important economic crop in tropical and subtropical regions, suffers from chilling injuries caused by postharvest low-temperature storage, which seriously affect its quality and economic benefits. Traditional detection methods have limitations such as low efficiency and strong destructiveness. This study designs and implements a flexible visible light spectral sensing system based on visible light spectral sensing technology and low-cost environmentally friendly flexible circuit technology. The system is structured based on a perception-analysis-warning-processing framework, utilizing laser-induced graphene electroplated copper integrated with laser etching technology for hardware fabrication, and developing corresponding data acquisition and processing functionalities. Taking Yunnan Yumang as the research object, a three-level chilling injury label dataset was established. After Z-Score standardization processing, the prediction accuracy of the SVM (Support Vector Machine) model reached 95.5%. The system has a power consumption of 230 mW at 4.5 V power supply, a battery life of more than 130 days, stable signal transmission, and a monitoring interface integrating multiple functions, which can provide real-time warning and intervention, thus offering an efficient and intelligent solution for chilling injury monitoring in mango cold chain storage. Full article

The characterization of airborne submicrometric composite structures’ debris is a challenge in the field of environmental monitoring and control. The work presented here aims to develop a new quantitative method to measure elemental mass concentrations via particle sampling and Transmission Electron Microscopy—Energy-Dispersive X-ray Spectroscopy (TEM-EDS). The principle is to collect airborne particles on a porous TEM grid, then add a certain mass of reference particles, and compare the relative mass percentages of elements from reference and sample particles via EDS. Diverse pairs of airborne particles (RbCl, CsCl, NaCl, SrCl₂, Ga(NO₃)₃, braking particles) were deposited on one TEM grid, and the experimental elemental mass ratios were measured by EDS and compared with the theoretical values. Results show that the quantitative and homogeneous collection of reference particles, such as RbCl, on the TEM grid could be suitable. For all the tested conditions, the absolute deviations between the theoretical elemental mass ratios and the experimental ratios remain lower than 8%. Thus, the mass concentration of Fe from the braking aerosol is calculated as 107 µg/m³. Compared to the cumbersome real-time instrument, this new method for mass characterization appears to be convenient, and requires a short time of aerosol sampling at the workplace. This approach ensures safety and practicability when assessing, e.g., the exposure risk of hazardous materials. Full article

With the rapid advancement of urbanization, urban transportation systems are facing increasingly severe congestion challenges, especially at traditional roundabouts. The rapid increase in vehicles has led to a sharp increase in pressure at roundabouts. In order to alleviate the traffic pressure in the roundabout, this paper changes the road design parameters of the roundabout, uses a CFD method combined with sensitivity analysis to study the influence of different inlet angles, lane numbers, and the outer radius on the pressure, and seeks the road design parameter scheme with the optimal mitigation effect. Firstly, the full factorial experimental design method is used to select the sample points in the design sample space, and the response values of each sample matrix are obtained by CFD. Secondly, the response surface model between the road design parameters of the roundabout and the pressure in the ring is constructed. The single-factor analysis method and the multi-factor analysis method are used to analyze the influence of the road parameters on the pressure of each feature point, and then the moment-independent sensitivity analysis method based on the response surface model is used to solve the sensitivity distribution characteristics of the road design parameters of the roundabout. Finally, the Kriging surrogate model is constructed, and the NSGA-II is used to solve the multi-objective optimization problem to obtain the optimal solution set of road parameters. The results show that there are significant differences in the mechanism of action of different road geometric parameters on the pressure of each feature point of the roundabout, and it shows obvious spatial heterogeneity of parameter sensitivity. The pressure changes in the two feature points at the entrance conflict area and the inner ring weaving area are significantly correlated with the lane number parameters. There is a strong coupling relationship between the pressure of the maximum pressure extreme point in the ring and the radius parameters of the outer ring. According to the optimal scheme of road parameters, that is, when the parameter set (inlet angle/°, number of lanes, outer radius/m) meets (35.4, 5, 65), the pressures of the feature points decrease by 34.1%, 38.3%, and 20.7%, respectively, which has a significant effect on alleviating the pressure in the intersection. This study optimizes the geometric parameters of roundabouts through multidisciplinary methods, provides a data-driven congestion reduction strategy for the urban sustainable development framework, and significantly improves road traffic efficiency, which is crucial for building an efficient traffic network and promoting urban sustainable development. Full article

A novel forming process was developed to create rotating elements in collector coins by shaping thin-walled tubes into toroidal shells in a single stage. This process introduces new aesthetic effects and functionalities to the coin market. By controlling the end-forming of a tube, a gap is created between the toroidal shell and the coin blank, allowing free rotation. The technique, a modification of conventional tube end inversion, results in distinct deformation mechanics. Through finite element modeling and experimentation, key process variables were identified, ensuring sound hollow toroidal shells with efficient rotation. Full article

Breast cancer represents a significant public health concern in both Western countries and Asia. Accurate and early detection is critical to improving long-term patient survival. For physicians to understand the classification and decision rules and to evaluate their results, it is preferable to use white box approaches to develop prediction models. This paper proposes a novel classification technique for extracting malignant prediction rules from training datasets containing numerical and binary nominal attributes. The classification technique introduced in this study facilitates the discovery of breast cancer patterns by integrating a real-coded genetic algorithm, an adaptive directed mutation operator, and a two-level malignant-rule-mining process. The experimental results, compared with existing rule-based methods from previous studies, demonstrate that the proposed approach generates simple and interpretable decision rules and effectively identifies patterns that lead to accurate breast cancer classification. Full article

Camera-centric perception has matured into a cornerstone of modern autonomy, from self-driving cars and factory cobots to underwater and planetary exploration. This review synthesizes more than a decade of progress in vision-based robotic navigation through an engineering lens, charting the full pipeline from sensing to deployment. We first examine the expanding sensor palette—monocular and multi-camera rigs, stereo and RGB-D devices, LiDAR–camera hybrids, event cameras, and infrared systems—highlighting the complementary operating envelopes and the rise of learning-based depth inference. The advances in visual localization and mapping are then analyzed, contrasting sparse and dense SLAM approaches, as well as monocular, stereo, and visual–inertial formulations. Additional topics include loop closure, semantic mapping, and LiDAR–visual–inertial fusion, which enables drift-free operation in dynamic environments. Building on these foundations, we review the navigation and control strategies, spanning classical planning, reinforcement and imitation learning, hybrid topological–metric memories, and emerging visual language guidance. Application case studies—autonomous driving, industrial manipulation, autonomous underwater vehicles, planetary rovers, aerial drones, and humanoids—demonstrate how tailored sensor suites and algorithms meet domain-specific constraints. Finally, the future research trajectories are distilled: generative AI for synthetic training data and scene completion; high-density 3D perception with solid-state LiDAR and neural implicit representations; event-based vision for ultra-fast control; and human-centric autonomy in next-generation robots. By providing a unified taxonomy, a comparative analysis, and engineering guidelines, this review aims to inform researchers and practitioners designing robust, scalable, vision-driven robotic systems. Full article

The mechanisms of in-service damage caused to the cylinder units of internal combustion engines (ICE) during their operation are analyzed. Long-term operation under harsh conditions, failure to comply with operating conditions, and breach of design and technology standards were found to be the major reasons for the initiation and propagation of in-service defects. The life of ICE cylinder liners is proposed to be extended by forming regular microreliefs. This represents a promising surface engineering strategy. Axial lines of the regular microrelief’s grooves were considered using analytical dependencies, which helped determine their coordinates and those of their equidistant. The authors simulated the pattern according to which the groove axes of type II regular microrelief could be aligned on the inner surface of the cylinder liner. To this end, a tool with three deforming elements was used. Technical means have been developed to implement this technology on the working surfaces of the liner–piston group’s mating parts. Full article

Accurate prediction of ship fuel consumption is essential for improving energy efficiency, optimizing mission planning, and ensuring operational integrity at sea. However, during complex tasks such as high-speed maneuvers, fuel consumption exhibits complex dynamics characterized by the coexistence of baseline drift and transient peaks that conventional models often fail to capture accurately, particularly the abrupt peaks. In this study, a hybrid prediction model, DGM-SVR, is presented, combining a rolling dynamic grey model (DGM (1,1)) with support vector regression (SVR). The DGM (1,1) adapts to the dynamic fuel consumption baseline and trends via a rolling window mechanism, while the SVR learns and predicts the residual sequence generated by the DGM, specifically addressing the high-amplitude fuel spikes triggered by maneuvers. Validated on a simulated dataset reflecting typical fuel spike characteristics during high-speed maneuvers, the DGM-SVR model demonstrated superior overall prediction accuracy (MAPE and RMSE) compared to standalone DGM (1,1), moving average (MA), and SVR models. Notably, DGM-SVR reduced the test set’s MAPE and RMSE by approximately 21% and 34%, respectively, relative to the next-best DGM model, and significantly improved the predictive accuracy, magnitude, and responsiveness in predicting fuel consumption spikes. The findings indicate that the DGM-SVR hybrid strategy effectively fuses DGM’s trend-fitting strength with SVR’s proficiency in capturing spikes from the residual sequence, offering a more reliable and precise method for dynamic ship fuel consumption forecasting, with considerable potential for ship energy efficiency management and intelligent operational support. This study lays a foundation for future validation on real-world operational data. Full article

Background: Electroencephalography (EEG) signals play a crucial role in diagnosing epilepsy by reflecting distinct patterns associated with normal brain activity, ictal (seizure) states, and interictal (between-seizure) periods. However, the manual classification of these patterns is labor-intensive, time-consuming, and depends heavily on specialized expertise. While deep learning methods have shown promise, many current models suffer from limitations such as excessive complexity, high computational demands, and insufficient generalizability. Developing lightweight and accurate models for real-time epilepsy detection remains a key challenge. Methods: This study proposes a novel dual-channel deep learning model to classify epileptic EEG signals into three categories: normal, ictal, and interictal states. Channel 1 integrates a bidirectional long short-term memory (BiLSTM) network with a Squeeze-and-Excitation (SE) ResNet attention module to dynamically emphasize critical feature channels. Channel 2 employs a dual-branch convolutional neural network (CNN) to extract deeper and distinct features. The model’s performance was evaluated on the publicly available Bonn EEG dataset. Results: The proposed model achieved an outstanding accuracy of 98.57%. The dual-channel structure improved specificity to 99.43%, while the dual-branch CNN boosted sensitivity by 5.12%. Components such as SE-ResNet attention modules contributed 4.29% to the accuracy improvement, and BiLSTM further enhanced specificity by 1.62%. Ablation studies validated the significance of each module. Conclusions: By leveraging a lightweight design and attention-based mechanisms, the dual-channel model offers high diagnostic precision while maintaining computational efficiency. Its applicability to real-time automated diagnosis positions it as a promising tool for clinical deployment across diverse patient populations. Full article

Nanomaterials are present everywhere today and represent the new industrial revolution. Depending on the application, there are many ways to synthesize nanomaterials with different properties. The industrial production of nanomaterials faces various challenges at different stages, going from conception and design to implementation and scaling-up of the production process, which can limit the growth of practical application at a large-scale scope, such as due to the lack of reproducibility, safety, and environmental impact. Here, we discuss current advances achieved for nanomaterial production at a large scale, encompassing a range of synthetic strategies and post-treatment modifications used to enhance the nanomaterials’ performance. A particular interest is devoted to highlighting the progress of MoS₂ nanomaterials’ application. Thus, overcoming those discussed challenges becomes a new prospect for the future perspectives of industrial nanomaterials and nanotechnologies. Full article

Since the beginning of the last decade, digital technological achievements have ushered the economies of developed countries into the fourth industrial revolution, transforming industries into smart ones, referred to as “Industry 4.0”, enabling them to innovate as a prerequisite for sustainable development and economic growth. At the same time, the European Union’s institutions are adopting strategies and programs to transform the European industry into a climate-neutral one, aiming to achieve this by 2050. The authors, participating in the introduction of Lean tools and digital technologies into one of the European shipyards using the “CULIS” (Connect Universal Lean Improvement System) methodology, recognize the high potential of its contribution to the European Commission’s guidelines for transitioning the economy to a sustainable one, and for this purpose, they present it in this paper. Namely, the methodology in question not only theoretically results in a “quick” implementation of tools and doctrines—with an approximately 36-month total duration of the process—but also encompasses as many as three transformations: Lean, digital, and green; an analysis of a methodology with such characteristics significantly adds to the originality of this study. The current stage of the observed shipyard’s “triple” transformation process already results in significant improvements—e.g., an increase in productivity by around 21% or a reduction in sales process costs by 38%. However, given its ongoing pilot phase, (further) analyses of improvements in (European) shipbuilding competitiveness and profitability that can be achieved through digital Lean management of projects’ realization process are implied. Full article

Modern fighter aircraft have an increasing need for at least a moderate level of stealth, and the shape design must bear a part of this constraint. However, the high frequency of close range radar makes high-fidelity radar cross-section computation methods such as the boundary element method too expensive to use in a gradient-free optimization process. On the other hand, asymptotic methods are not able to accurately predict the RCS of complex shapes such as intake cavities. Hence, the need arises for efficient and accurate methods to compute the gradient of high-fidelity radar cross-section computation methods with respect to shape parameters. In this paper, we propose an adjoint formulation for the boundary element method to efficiently compute these gradients. We present a novel approach to calculate the gradient of high-fidelity radar cross-section computations using the boundary element method. Our method employs an adjoint formulation that allows for the efficient computation of these gradients. This is particularly beneficial in shape optimization problems where accurate and efficient methods are crucial to designing modern fighter aircraft with stealth capabilities. By avoiding the need for solving the actual adjoint problem in certain cases, our formulation provides a more streamlined solution while still maintaining high accuracy. We demonstrate the effectiveness of our method by performing shape optimization on various shapes, including simple geometries like spheres and ellipsoids, as well as complex aircraft shapes with multiple variables. Full article

This work will present an algorithmic approach for robust control focusing on hydraulic–mechanical systems. The approach is applied to a hydraulic actuator driving a cart with an inverted pendulum. The algorithmic approach aims to satisfy two robust control requirements for single input single output (SISO) linear systems with nonlinear uncertain structure. The first control requirement is robust stabilization, and the second is robust asymptotic command following for arbitrary reference signals. The approach is analyzed in two stages. In the first stage, the stability regions of the controller parameters are identified. In the second stage, a Particle Swarm Optimization Algorithm (PSO) is applied to find suboptimal solutions for the controller parameters in these regions, with respect to a suitable performance cost function. The application of the approach to a hydraulic actuator, driving a cart with an inverted pendulum, satisfies the goal of achieving precise control of the pendulum angle, despite the system’s inherent physical uncertainties. Full article