This study presents a digital twin of 5-axis milling processes for thin-walled parts to predict and prevent the occurrence of static deflection issues, surface location error due to forced vibration, and regenerative chatter. Driven by real-time machine signals or machining programs, the twin continuously simulates the milling process given an initial workpiece and any 5- axis toolpath. Thanks to the implementation of an automatic mesher and a finite element model generator, the stiffness and dynamics of the part are determined based on its current geometry. This enables the digital twin to predict and visualize in 3D the static deflection, forced vibration, and process stability, allowing optimization of the process and avoidance of these issues. Finally, the twin is validated on a 5-axis machining case of a steel blade, in which a surface location error caused by a resonance is diagnosed and eliminated.
Induction motors are the most widely used type of motor in industry. They are designed in single-phase and three-phase versions. Many critical parameters are involved in the design of an induction motor, and each parameter has a specific impact on motor performance. One of the most important design parameters is the geometry of the stator and rotor slots. In this study, motor design was carried out using 10 different optimization algorithms for slot geometries. Furthermore, iron losses and efficiency, which are indicators of motor performance, were optimized using the slot geometry designs. Genetic Algorithm, Tabu Search Algorithm, Simulated Annealing Algorithm, Whale Optimization Algorithm (WOA), and their variants were used for optimization. The objective function was to minimize iron losses and maximize motor efficiency by optimizing the stator and rotor slot dimensions for energy saving. The main motivation for choosing the WOA method was its better convergence speed and solution accuracy for nonlinear problems such as induction motor design. WOA and its variants were applied to this problem for the first time, and iron losses and efficiency were optimized. The results showed that WOA and its variants achieved more successful results compared with the other optimization methods. As a result of the optimization, iron losses were reduced by 41.55%, leading to a 0.802% increase in the efficiency of the induction motor.
Risk factors are highly coupled and continuously evolve throughout the entire power extension operation process. Existing risk management methods, primarily based on static experience, are insufficient to support real-time early warning and effective control. To address this issue, this paper uses multi-source time-series data from the entire power extension process as input. First, a Dynamic Bayesian Network (DBN) is constructed to describe the causal relationships and stage evolution characteristics of risks, enabling online updates of risk states within a probabilistic inference framework. Subsequently, a Long Short-Term Memory (LSTM) network is used to deeply model the risk-related time-series characteristics and predict potential risk evolution trends. By fusing the risk state probabilities output by the DBN with the LSTM prediction results, a unified dynamic risk early warning index is formed, and targeted control decision support is achieved based on the reverse inference of key risk nodes. Experimental results show that the average early warning lead time of the DBN+LSTM fusion model reaches 18.7 minutes; the risk state identification accuracy during the power extension stage is 86.8%; the overall false alarm rate and missed alarm rate are 5.21% and 1.82%, respectively. The research outcomes demonstrate that this method provides an effective technical approach for dynamic risk early warning and control throughout the entire process of power grid expansion, combining mechanistic explanation and data-driven capabilities. Keywords: Dynamic Risk Prediction; Electric Connection Lifecycle; Dynamic Bayesian Network; Long Short-Term Memory; Risk Control Strategy
Artificial intelligence (AI) advancements within logistics receive significant academic attention; however, existing research often overlooks innovation decision regarding delivery technology, particularly the mechanism of user influence. An evolutionary game model incorporating logistics enterprises, R&D institutions, and consumers was established from the perspective of users in this study. The model analyzed the strategic choices of these three actors in AI delivery innovation and examines the factors governing their decisions. Furthermore, simulation analysis verified the evolutionary equilibrium and stability of these strategies. Results revealed that: (1) Enhancing collaborative value-added effect amplifies tripartite benefits, thereby accelerating the transition to an optimal equilibrium of “introduction, technological innovation, active participation”. (2) Reducing implementation costs encourages logistics enterprises to adopt new AI technologies proactively. (3) Effective cost control further sustains the commitment of R&D institutions to achieving technological breakthroughs. (4) Increased benefits for consumers significantly bolster participation, which subsequently incentivizes enterprises and R&D institutions to pursue active innovation strategies. This study offers a new research perspective for AI delivery innovation and provides a feasible management framework for logistics stakeholders, fostering sustainable industry development. Keywords: Artificial intelligence technology; Logistics delivery; Evolutionary game; Logistics enterprise; User
Abstract- Phishing effects have turned to be a most pre-dominant attacks encountered by Internet users especially for users in IoT environment. Many investigators attempt to provide solution which leads to a major prediction disaster. This work proposes a novel Machine Learning (ML) approach to handle Phishing effect on Email over the targeted sectors. Thus, effectual phishing detection is required to needed to trace the phishing effect over the email with a periodic alarm rate. This work proposes a novel regression-based linear support vector model (r-lSV) to mitigate the phishing effect problem and provides awareness by analyzing the phishing features to predict and prevent the phishing scams in its earlier stage. The proposed prediction model partitions the dataset into testing and training to analyze the inherent phishing characteristics over email. The proposed model partitions the phishing and non-phishing using online available dataset. The functionality of the anticipated model is compared with other machine learning approaches. The functionality of the proposed model is optimized using Bald Eagle Eye Optimizer (BEO) to attain global outcomes. The proposed model intends to give superior outcomes based the feature learning significance and classification. The proposed r-lSV establishes better trade-off compared to other approaches. Keywords- phishing effect, prediction, machine learning, feature representation, scams
This paper presents a controlled comparative study of traditional machine learning algorithms for thematic text classification, focusing on the impact of preprocessing strategies and class imbalance on model performance under different data conditions. Two experimental scenarios were considered: the Women’s Clothing E-Commerce Reviews dataset, characterized by a highly imbalanced class distribution, and the AG News dataset, used as a balanced reference dataset. The preprocessing pipeline included text cleaning, stopword removal, tokenization, lemmatization, TF-IDF vectorization, class balancing, and dimensionality reduction. Six models were evaluated: Naive Bayes, Random Forest, Support Vector Machine, Logistic Regression, K-Nearest Neighbors, and K-Means as an unsupervised baseline. The results show that Random Forest achieved the best performance on the imbalanced Clothing dataset, reaching an accuracy of 0.763, an F1-score of 0.777, and an MCC of 0.739. In contrast, K-Nearest Neighbors obtained the highest scores on the balanced AG News dataset, with an accuracy and F1-score of 0.927. Naive Bayes showed limitations when dealing with sparse and ambiguous textual representations. The main contribution of this work is a systematic experimental analysis that highlights how preprocessing decisions, feature sparsity, and class distribution influence the behavior of traditional machine learning models in text classification tasks. These findings provide useful insights for model selection and evaluation under different data scenarios. Key Words: Text Classification; Text Preprocessing; Machine Learning; Text Analysis; Class Imbalance
Multi-stage manufacturing systems require the highest possible level of availability. Due to the large batch size required for proper manufacturing, it is necessary to certify failure-free manufacturing time. Specific multi-stage machines must be characterized to calculate availability. Furthermore, preventive maintenance has been studied in several cases, thus avoiding real non-production costs due to component failures. OEE and TPM standards allow to study this problem at single-stage machines. This article presents two Novel methods for decision-making and studying the effect of unexpected failures in multi-stage machines on availability optimization and penalty costs. The first explores the real penalty cost, and the second, the impact on availability. Both methods require prior characterization of the multi-stage machine and time-dependent condition adjustment of all components. A real-life case is presented to test both methods. The first allows for a more in-depth risk analysis of specific components and enables decision-making before manufacturing systems are designed by leveraging the Internet of Things and secure remote virtual access to multi-stage manufacturing systems. In the second case, the result in terms of availability appears to be lower, but the economic improvement significantly influences the end user's decision-making. The study link to OEE standards to fix a point for further researches. Keywords: Decision Making, Penalty cost, Maintenance, Availability, Multistage Machine.
This research presents a novel continuous controlled phase shifter network that uses an adjustable phase shifter known as the Dual Band Phase Shifter Model (DBPSM) with Antenna Merged Waveguide (AMW) which has a phase tuning range of 90? and the other two of 180? for adjusting the progressive phase difference at the matrix network's output ports. The phase difference in the 90? range can be tailored for each input stimulus when all phase shifters are simultaneously controlled using two-channel digital voltages. Only four input ports must be switched to achieve the progressive phase difference over its full 360? range. Researchers conduct a theoretical analysis of the scenario and develop closed-form equations that minimize the number of phase shifters and their phase tuning ranges, thereby reducing the control complexity. The design theory is verified by producing an iteration of the suggested tunable AWM at 5.8 GHz. The results of the experiment and simulation agree on most points. It is possible to guide the observed radiation phase between -51? and 52? by combining two control voltage adjustments with four patch antenna elements. Keywords: Phase shifter-relaxed feeding network, waveguide, tunable phase difference, control-relaxed steering
This structured literature review evaluates the climate-dependent energy performance of smart glazing technologies, electrochromic (EC), thermochromic (TC), and photochromic (PC), drawing on twenty peer-reviewed studies selected through a PRISMA-based methodology. The literature reviewed indicates that EC and TC systems appear to reduce cooling loads more effectively in warm climates, while PC glass performs better in cold climates by preserving daylight and reducing heating demand. Performance varies with façade orientation, control strategy, and solar radiation intensity. The findings suggest a comparative technical framework linking smart glazing type, climatic zone, and building typology. This review provides insights into façade design and material selection for energy-efficient and climate-responsive buildings. Key Words: smart glass; energy efficiency; climate-responsive design; electrochromic glass; photochromic glass; building envelopes
The widespread application of artificial intelligence technology in logistics delivery has made technological innovation the core driver of industry development. Given that existing evolutionary game studies rarely address collaborative innovation mechanisms in logistics scenarios, this study constructed a tripartite evolutionary game model involving “government-logistics enterprises-R&D institutions.” It delved into the strategic evolution patterns and key influencing factors among these three entities in AI technological innovation for logistics delivery. The findings reveal: (1) The system exhibits multiple evolutionarily stable equilibrium points, with (government incentives, logistics enterprises introduction, R&D institutions technological innovation) representing the ideal equilibrium state. The system demonstrates strong convergence properties, ultimately converging toward the optimal collaborative state. (2) Factors such as innovation costs, introduction costs, and government subsidies significantly influence the evolutionary process. Increased government subsidies to logistics enterprises and R&D institutions accelerate convergence toward the ideal equilibrium, while higher innovation costs and introduction costs slow convergence. (3) The probability of initial strategies adopted by actors affects evolutionary efficiency. Higher initial cooperation willingness among the government, logistics enterprises, and R&D institutions accelerates the system’s convergence to the ideal equilibrium. This study enriches the research perspective on AI-driven logistics delivery technology innovation and provides valuable management insights for government formulation of collaborative innovation policies, as well as cooperation between logistics enterprises and R&D institutions. Keywords: Artificial intelligence technology, Logistics delivery, Evolutionary game theory, Technological innovation, Collaborative mechanisms
Urban water supply management requires ensuring continuous water availability; however, increasing drought events necessitate prioritizing uses and implementing emergency plans. The obligation to develop such plans was established in 2001 under Law 10/2001 of the National Hydrological Plan, within the framework of the European Water Framework Directive 2000/60/EC. Severe droughts during the second half of the 20th century led to the development of specific guidelines and protocols, culminating in the nationwide “Guide for the Preparation of Drought Emergency Plans in Urban Water Supply Systems.” With technological progress, traditional restrictions and supply cuts have been progressively replaced by more sustainable strategies such as demand management through incentives, water reuse, and the creation of strategic reserves. Reclaimed water—particularly through indirect reuse—has become a strategic resource. However, its direct use remains limited by national regulations (Royal Decree 1085/2024) and the European framework concerning public health safety and environmental protection, which conditions its direct application in urban uses and affects associated costs. Desalination, in turn, is essential in regions with low rainfall, as it represents a climate-independent resource, although it entails significant energy and environmental challenges. The Greater Bilbao case illustrates how the integration of reservoirs, strategic conveyance systems, and emergency infrastructures can enhance supply resilience. Overall, modern water management requires resource diversification—including reclaimed, desalinated, and conventional surface water sources—and the consolidation of flexible infrastructures to address increasingly severe drought scenarios.
Future power grids will be characterized by the extensive integration of distributed generation (DG) from renewable energy sources, introducing new challenges for system operation and management. This study addresses the underutilized reactive power potential of distributed generation (DG) in power systems through a comprehensive five-objective optimization framework. The framework simultaneously minimizes transmission losses, voltage deviations, generation costs, voltage stability indices, and DG reactive power requirements. Four metaheuristic algorithms (SPEA2, MOPSO, NSGA-II, and MOEA/D) are comparatively evaluated on a modified IEEE 57-bus system with solar DG capacity across five operational scenarios over 24 hours. Results show that coordinated DG reactive support reduces transmission losses compared to active-power-only operation. SPEA2 and MOPSO provide better Pareto front quality with higher hypervolume values and DG units can provide 20-40 MVAr of reactive support, representing significant economic value under emerging reactive power markets. These findings provide practical guidelines for system operators designing reactive power strategies in networks with high renewable penetration, showing that proper DG coordination improves technical performance while creating new revenue opportunities through ancillary services.
Researchers have put forward that the variations in the climate based on various weather conditions directly affect wind power forecasting. Predicting weather changes and wind power output accurately and theoretically using statistical prediction models is complex. With conventional learning models, forecasting long-term wind power can be made to work with mean absolute percentage error of ten per cent to seventeen per cent; this did not meet our renewable energy project's engineering requirements. The Generative Convolutional Information Encoding Network model (GCIENM) is proposed to achieve the correlations among power generation and meteorological parameters. In the wind power forecasting field, the presented technique has broad applicability. Henceforth, the research study focuses on the long-term, one day to three days ahead in wind power prediction with the (MAPE) of below 10% by employing GCIENM-based ATM and MVPNN. When we experimented, the GCIENM model performed better using the outputs of wind power generation in a wind power plant located in Scada and other historical weather data. The production of the experiment shows a MAPE value of 6% in the prediction of wind power in three days; for our project, this Value is sufficient for our requirement. This work finally compared the performance based on the proposed prediction model's performance for power forecasting. Our experiment showed that the GCIENM performs better than the three other models for predicting wind power by the parameters of forecast accuracy, error reduction stability and data input volume. Keywords- renewable energy, prediction, deep learning, error rate, actual prediction
This study presents the design, implementation and validation of an experimental quality control cell for the automatic inspection of defects in machined parts using an integrated machine vision system. The cell incorporates commercially available hardware, real-time image processing algorithms, and rigorous statistical analysis. In addition, the capacity for traceability and automatic reporting is highlighted, facilitating real-time decision-making and predictive maintenance planning. The system was configured to perform in-line inspections, detecting surface and dimensional defects in real-time. The methodology included the design of the inspection system, the programming of the camera and its integration into the control system. The implementation of the IVG300 camera proved to be an efficient, scalable solution aligned with the principles of Industry 4.0. This study concludes that the incorporation of intelligent vision systems is a key strategy to increase competitiveness in advanced manufacturing environments. Key Words: machine vision, quality control, defects, automatic inspection, industry 4.0
The problem of osteoporosis is an important concern facing our society today which requires methods for detecting it in its early stages that do not expose individuals to radiation, such as those utilized by dual-energy X-ray absorptiometry (DXA) which is currently considered standard practice for measuring bone mineral density (BMD). The high cost and limited number of DXA machines prohibit many people from being screened for osteoporosis. To address this limitation, Bone Radar, an electro-acoustic device designed for cost-effective and portable osteoporosis screening, was developed. The system operates by transmitting controlled acoustic signals into biological tissue and detecting the returning echoes. These echoes are subsequently analyzed using spectral signal processing techniques. Calibration experiments conducted on materials with known densities demonstrated a measurable correlation between the power (intensity) and the density of the tested material. Preliminary evaluations were then performed to assess the device’s ability to differentiate between individuals with normal BMD and one diagnosed with osteoporosis. The results indicate that Bone Radar successfully distinguished between the two groups based on their acoustic responses. These findings suggest that Bone Radar may provide a promising non-invasive and radiation-free approach for the early screening of osteoporosis.
Autonomous vehicles must sense their surroundings and exchange data reliably in dense traffic and poor visibility. Many millimetre-wave ISAC proposals assume high-rate converters, near-ideal RF chains, and large compute budgets that do not match production constraints. This paper presents a low-rate ISAC design that co-designs waveform, baseband processing, and scheduling around converter and ECU limits, while using roadside units (RSUs) as cooperative illuminators and relays. The design combines a hybrid FMCW–OFDM frame with impairment-aware estimation that explicitly accounts for phase noise, beam squint, mutual coupling, quantisation, and calibration drift. A cross-layer scheduler enforces an explicit ADC bit-budget per frame while meeting safety-critical latency and reliability targets, and it reuses RSU beacons for bistatic sensing and congestion-aware allocations at intersections. We evaluate the full stack in an urban four-way intersection scenario using simulation with hardware-calibrated impairment models. Relative to an 8-bit full-rate vehicle-only baseline, the RSU-assisted 3-bit low-rate configuration achieves 58% ADC-bit savings while maintaining 0.21 m range RMSE and 6.98 ms 99th-percentile one-way latency under the URLLC constraint set. These results indicate a practical path to reduce cost and power while preserving perception quality and V2X service in RSU-covered urban areas.
The present article analyzes the modernization of emergency lighting systems in industrial environments through the adoption of the IEC 62034 standard, which regulates automatic test systems (ATS). The study compares two technological approaches: systems with individual batteries and systems with centralized batteries, both evaluated with and without automated testing. Key aspects such as operational efficiency, safety, regulatory compliance, and maintenance cost reduction are examined. The analysis is framed within the common challenges faced by industries, such as the need to upgrade infrastructure, improve sustainability, and ensure the reliability of evacuation systems. Through a real-world case study, the limitations of traditional systems—such as low reliability and high operational costs—are identified, and key performance indicators are compared before and after the implementation of automated solutions, including preventive maintenance, energy efficiency, and overall system reliability. The results show that systems with automatic testing offer greater economic profitability and faster return on investment. In the analyzed case, luminaires with individual batteries and automatic testing proved to be the most cost-effective option. However, centralized battery solutions stand out for their lower environmental impact and their suitability in facilities with high-mounted luminaires, where maintenance is more complex. This study provides a practical guide for companies seeking to optimize their emergency lighting systems in response to the demands of an increasingly competitive and regulated industrial environment. Keywords: emergency lighting, illumination, IEC 62034, evacuation, maintenance.
