https://ijrtei.sbu.ac.ir/ International Journal of Research and Technology in Electrical Industry en daily 1 Tue, 01 Jul 2025 00:00:00 +0330 Tue, 01 Jul 2025 00:00:00 +0330 https://ijrtei.sbu.ac.ir/article_106456.html In this paper, a novel approach based on a non-monotonic Lyapunov–Krasovskii functional is first proposed for the comprehensive stability analysis of discrete-time switched systems with time delays in an m-step-ahead scheme. Subsequently, by extending these results, non-uniform robust stability conditions suitable for systems with uncertainties are derived. In this method, the stringent requirement of the uniform decrease of the Lyapunov–Krasovskii functional is replaced by non-monotonic conditions; that is, the functional is allowed to increase at certain steps, while its overall trend must remain decreasing. Consequently, this approach accommodates a broader class of functionals for stability analysis. Within the non-monotonic Lyapunov–Krasovskii framework, a set of sufficient conditions in the form of linear matrix inequalities (LMIs) is formulated to assess global asymptotic stability of discrete-time delayed systems. Moreover, Abel’s lemma is employed to further reduce conservatism in the stability analysis of switched systems compared to previous studies. Unlike its continuous-time counterpart, the discrete-time model exhibits higher complexity due to the interactions among subsystems induced by switching. To demonstrate the effectiveness of the proposed method, simulation results are provided for two numerical examples. https://ijrtei.sbu.ac.ir/article_106597.html The structure and ordering of nodes in power grids significantly affect the performance of algorithms used for simulation, optimization, and visualization. While synthetic generation of power grids has gained momentum in both transmission and distribution domains, limited research has addressed how node indexing impacts matrix sparsity patterns and computational performance. This paper presents a comprehensive study of node reordering techniques aimed at achieving consecutive numbering in complex power networks. Four categories of methods are explored: heuristic approaches, optimization-based strategies, graph-theoretic algorithms, and artificial intelligence (AI)-driven models. We revisit existing heuristics and propose a novel metaheuristic optimization method that maximizes diagonal density across multiple bandwidth levels. For radial distribution grids, we develop a graph-based method using depth-first search (DFS) that aligns node numbering with physical structure. In addition, we evaluate the feasibility of using convolutional and feedforward neural networks to learn reordering patterns from data. Despite training on thousands of synthetic graphs, standard AI models fail to produce valid permutations, highlighting the need for permutation-aware architectures. Extensive experiments on synthetic transmission and distribution networks confirm that method selection should align with network topology. Optimization excels in meshed systems, while graph-based DFS is ideal for tree-like networks. A comparative analysis is provided to guide future applications and research in this space. https://ijrtei.sbu.ac.ir/article_106598.html Contemporary steel manufacturing is undergoing a strategic transformation through Industry 4.0 adoption, leveraging artificial intelligence (AI), automated systems, and advanced analytics to achieve dual objectives of operational efficiency and environmental sustainability. This study presents an AI-integrated vocational education and reskilling framework specifically designed for steel production environments, synchronizing workforce development with technological innovation. Developed using operational data from an active direct-reduction facility, this framework integrates two complementary components: machine learning (ML) algorithms optimizing metallization parameters and predictive maintenance protocols, paired with an adaptive training curriculum utilizing digital twin simulations and generative AI for skill personalization. Result outcomes demonstrate that AI-driven training significantly boosts workforce skills, while ML improves metallization efficiency and reduces gas consumption. The work highlights the value of aligning vocational education and smart manufacturing infrastructure, revealing measurable improvements in both production metrics and workforce agility. The demonstrated approach provides a replicable blueprint for industrial upskilling, positioning AI-curriculum integration as a strategic imperative for maintaining sector competitiveness while advancing circular production paradigms. https://ijrtei.sbu.ac.ir/article_106599.html Open-circuit (OC) fault identification in Voltage-Source Inverters (VSIs) is a critical challenge for the reliability of power systems and motor drives. While Deep Learning (DL) technologies offer automatic feature extraction, they typically suffer from high computational costs and the requirement for massive labeled datasets, which are scarce in real-world industrial scenarios. This paper provides an innovative lightweight diagnostic approach utilizing Transfer Learning (TL) to address current issues of data scarcity and training inefficiency. The pre-trained SqueezeNet model, an efficient Convolutional Neural Network (CNN) with a lower parameter count, is fine-tuned to properly categorize various fault states. In the proposed methodology, the three-phase output current signals are first converted into time-frequency scalograms using the Continuous Wavelet Transform (CWT) to capture rich transient fault features. Subsequently, these visual representations are processed by the network. The proposed method achieves 99.90% accuracy on a small dataset (2000 samples) with considerably reduced training time relative to training deep models from scratch. These findings demonstrate the effectiveness and robustness of the suggested methodology for real-time fault diagnosis in inverters. https://ijrtei.sbu.ac.ir/article_106600.html Nudge theory, a concept in behavioral science, advocates for the use of reinforcement, encouragement, and subtle recommendations to encourage voluntary adherence and shape the motivations and decisions of individuals or groups. This approach has emerged as an influential method for steering consumer behavior in energy consumption, thereby optimizing energy system operations. In this regard, demand response (DR) policies in energy systems have significant potential to align with behavioral and nudge theory concepts. This paper incorporates positive real-world incentives into DR modeling for both electricity and heating systems, aiming to influence household and customer decision-making in energy consumption through behavioral concepts. Therefore, this study introduces an optimal scheduling framework for multi-carrier energy systems that incorporates nudge-based behavioral integrated demand response (NBIDR), which combines behavioral principles into a DR program (DRP) for both electricity and heating systems. The suggested mixed integer linear programming (MILP) framework was applied to the IEEE 33-bus test system. Simulation results demonstrate the effectiveness of the proposed framework by smoothing load profiles as well as decreasing operation costs and expected energy not supplied (EENS) for both electricity and heating infrastructures. https://ijrtei.sbu.ac.ir/article_106168.html The integration of gas and electricity networks is pivotal for efficient energy management, particularly with the rising penetration of renewable energy sources. Compressor stations are critical for maintaining gas pressure and flow, and their optimal operation can significantly enhance system flexibility and reduce costs. While previous studies have explored coordinated operation of compressor units, this research introduces a novel price-responsive coordination strategy for compressor stations comprising both gas-driven compressors (GDCs) and electric-driven compressors (EDCs) within an integrated gas and electricity network. The proposed strategy operates GDCs when gas prices are lower and EDCs when electricity prices are lower, aiming to optimize linepack storage in gas pipelines. Using a mixed-integer linear programming (MILP) model, we optimize the scheduling of compressors based on hourly energy prices while ensuring network constraints are met. Simulations on a 24-bus electricity and 19-node gas network over a 24-hour period demonstrate that this coordinated approach leads to a substantial increase in linepack storage and a reduction in operational costs compared to uncoordinated operation. Simulation results show that in the proposed design, i.e., by integrating the optimized linepack model into the compressor station model, the amount of carbon dioxide produced has decreased by 33.3% and the total operation costs have decreased by 1.57%. https://ijrtei.sbu.ac.ir/article_106741.html Microgrids operating in isolated mode often suffer from frequency instability due to unpredictable load variations and intermittent renewable energy generation. This paper proposes an enhanced load frequency control strategy that combines fuzzy logic with fractional-order PID dynamics to address these challenges. The fuzzy inference mechanism generates a nonlinear control action based on frequency deviation and its rate of change, enabling real-time, intelligent adaptation to disturbances. Fractional-order operators enhance damping and robustness across a wide operating range.Controller parameters are tuned using a Quadratic Inspired Optimization (QIO) algorithm to minimize performance indices like Integral Absolute Error (IAE) and Integral Time Absolute Error (ITAE). A key innovation is incorporating motor drive-based loads as controllable frequency-responsive elements, rather than mere disturbances. Despite their nonlinear nature, these loads are harnessed to boost dynamic performance and accelerate stabilization.MATLAB/Simulink simulations across scenarios—consumer loads only, renewables only, and combined—confirm the strategy reduces frequency deviations and improves transient performance over PID, FOPID, and fuzzy-based benchmarks, with superior error metrics and recovery times. This highlights the method's robustness for real-world microgrids. https://ijrtei.sbu.ac.ir/article_106742.html Non-isolated DC–DC converters offer an effective solution to the limitations of isolated counterparts. This study proposes a new non-isolated DC–DC converter that integrates a conventional boost converter with a diode–capacitor voltage multiplier cell (VMC). The higher voltage gain is achieved without increasing the number of inductors, which remains the same as in Cuk, SEPIC, and Zeta converters. Similar to boost, Cuk, and SEPIC converters, the proposed topology ensures continuous input current and subjects the input filter capacitor to low current stress. It also provides a common ground between the load and source while maintaining positive output polarity. In addition to delivering a high voltage gain, the converter achieves low and acceptable voltage/current stresses on the semiconductors. Both ideal and non-ideal operating modes are analyzed. Finally, experimental results are presented to validate the theoretical analysis and relation, with a 200 W prototype designed for 40 V input and 400 V output. https://ijrtei.sbu.ac.ir/article_106743.html Purpose: The rapid advancements in convolutional neural networks (CNNs) have significantly improved medical image analysis. The COVID-19 pandemic has impacted millions worldwide, with containment hindered by inadequate testing resources and inefficiencies in diagnostic methods. This study proposes and explores a novel framework employing an ensemble of advanced deep transfer learning techniques for accurate and consistent COVID-19 detection from Computed Tomography (CT) scans, reducing reliance on manual assessment.Materials and Methods: The proposed framework integrates standardized data pre-processing with fine-tuned heterogeneous transfer learning models, including CNN- and Transformer-based architectures. An ensemble learning strategy is implemented at the feature level using Principal Component Analysis (PCA) to fuse deep representations extracted from the most effective models. The framework has been evaluated on two large publicly available CT datasets, COVID-CT and SARS-CoV-2, comprising over 1,600 COVID-19 and 1,450 non-COVID-19 images.Results: Experiments have demonstrated that fusing five architectures—ResNet-50 v2, EfficientNet-B5, ViT, VGG16, and DenseNet-201—achieves superior diagnostic performance compared to individual models and existing frameworks, as reflected by improved F1-scores.Conclusion: The results have confirmed that integrating transfer learning with feature-level ensemble learning within a unified framework significantly enhances the robustness and accuracy of COVID-19 detection from CT images. The proposed methodology provides a scalable and reproducible solution that can be extended to other medical image-based diagnostic tasks. https://ijrtei.sbu.ac.ir/article_106783.html This paper introduces a compact, ultra-wideband Vivaldi antenna with parasitic elements, designed for 5G millimeter-wave (mmWave) applications. The antenna operates across a wide frequency range from 28.57 GHz to 80 GHz, covering essential high-frequency bands for current and future communication systems. The design features innovative elements such as curved circular flares and parasitic components, which boost both gain and directivity, thereby enhancing overall performance. The antenna is implemented on a Rogers RT5880 substrate with compact overall dimensions of 23 × 20 mm². Full-wave simulations using CST software show that the antenna achieves a peak gain of 10.5 dBi throughout the operating bandwidth and a radiation efficiency of up to 96%. To ensure impedance matching and reduce surface wave effects, a stepped-feedline is incorporated, promoting stable performance across the entire frequency range. The antenna exhibits desirable radiation characteristics, including low sidelobe levels and highly directional radiation patterns. With its compact size, high gain, and efficient operation, the antenna is well-suited for mmWave 5G communication systems, radar applications, and high-speed wireless networks. https://ijrtei.sbu.ac.ir/article_107024.html Hybrid AC/DC distribution networks (HDNs) have emerged as a promising architecture for integrating distributed energy resources and the growing number of DC loads. However, most existing expansion planning studies treat demand response (DR) as a homogeneous flexibility resource, neglecting the different flexibility characteristics of AC and DC loads. This simplification may lead to unrealistic estimations of demand-side flexibility and suboptimal planning decisions. This paper proposes an expansion planning framework for HDNs that incorporates differentiated DR modelling for AC and DC loads. In the proposed approach, distinct flexibility limits and participation costs are assigned to AC and DC loads to better represent their controllability characteristics. The planning problem is formulated as a mixed-integer nonlinear optimization model that determines network expansion decisions by incorporating DR utilization. The model is solved using a genetic algorithm implemented in MATLAB, while operational costs are evaluated through an optimal power flow module developed in GAMS. Three planning scenarios are analyzed to assess the effectiveness of the proposed framework: without DR, with uniform DR, and with differentiated DR for AC and DC loads. The results show that modelling differentiated DR improves demand-side participation and reduces the net present value of total planning costs compared with conventional uniform DR modelling. These findings highlight the importance of accurately representing heterogeneous demand flexibility in the planning of HDNs.