ECTI Transactions on Electrical Engineering, Electronics, and Communications

Multi Half Bridge LLC Resonant Inverter for Induction Hardening

Jirapong Jittakort — 2026-09-29

A multi half-bridge LLC resonant inverter with three induction coils is used to harden metal workpieces in this research. The system operates at 80–100 kHz and outputs 3 kW of power, based on to load variations. Power is controlled through changing the duty cycle of each inverter module's output voltage waveform and a phase-locked loop for frequency tracking. This method allows soft-switching and improves system efficiency. Stable operation and improved performance demonstrate that proposed system is suitable for induction hardening. The multi-inverter design improves power distribution, system dependability, flexibility, and scalability for many industrial applications.

A Systematic Design Approach of Non-Isolated LED Driver Based on Integrated Bridgeless Boost PFC Rectifier and ZVS Class-D Resonant Converter

Chainarin Ekkaravarodome — 2026-09-29

A systematic design approach for a high-power-factor integrated-stage, light-emitting diode (LED) driver for street-lighting applications is presented. The proposed scheme integrates a totem-pole bridgeless boost powerfactor correction (PFC) rectifier as a PFC stage and non-isolated zero-voltage switching Class-D resonant converter as a driver stage into an integrated stage by sharing power switches and a DC bus capacitor. High power factor is inherently achieved by operating the totem-pole bridgeless boost PFC rectifier in discontinuous conduction mode. The proposed systematic design and circuit operation are fully detailed using simplified equivalent circuits, which significantly reduces implementation complexity. This thorough methodology further provides a clear, step-by-step guide for component calculation, loss analysis, and component stress evaluation. The validity of this approach is confirmed by experimental results from a 100-W LED streetlighting driver prototype.

Improved Inductive WPT System Modeling with a ZVDS Class-DE Current-Driven Full-Bridge Rectifier for Accurate Conduction Loss Analysis

Chainarin Ekkaravarodome — 2026-09-29

This paper presents an improved inductive wireless power transfer (WPT) system model with a zero-voltage and zero-derivative-switching Class-DE current-driven full-bridge rectifier for accurate conduction loss analysis. Conventional models typically assume a near-infinite magnetizing inductance with a negligible magnetizing current, which fails in applications where a low magnetizing inductance induces a very high magnetizing current. This elevated magnetizing current causes a significant phase lag of the input current relative to the input voltage of the transmitter compensation network, leading to substantial reactive power and increased conduction losses. Accordingly, the proposed model includes the coupling coefficient and these magnetizing effects into the key equations. To validate the proposed model, the developed inductive WPT converter utilizing a phase-shift controller at an 88-kHz switching frequency has been experimentally evaluated under a 360-V input voltage and 350-V output voltage, covering an output power range from 0.2 kW to a rated power of 1.8 kW at a 50-mm transfer distance.

AI-Driven Conversational Interface for Enhancing Industrial Data Accessibility in Warehouse Management Systems

Thaninrath Witoontharaphong — 2026-09-29

AI-powered conversational interface designed to enhance industrial data accessibility in Warehouse Management Systems (WMS). The system addresses the challenge of enabling non-technical users to retrieve complex warehouse data without requiring SQL knowledge or database expertise, supporting data-driven decision-making in manufacturing operations. Our approach employs WangchanBERTa, a pre-trained Thai language model, for Natural Language Processing to interpret user intents and extract relevant entities from Thai questions. The system then converts natural language queries into SQL commands using Natural Language to SQL (NL2SQL) techniques to retrieve information from WMS databases. The proposed architecture consists of five key components, and the system's effectiveness is evaluated through implementation in a footwear manufacturing facility, where sample queries regarding inventory, stock levels, and operational metrics are tested.

Experimental results show the Intent Classification model achieves a Macro F1-Score of 87.84% and an Accuracy of 94.01%, demonstrating robust performance across multiple intent categories. System generates correct SQL queries for common warehouse queries, including product color, quantity, and stock availability. While certain complex intents exhibit lower accuracy due to contextual ambiguity and limited training data. This research provides a replicable framework to be extended to other industrial domains, thereby enhancing operational efficiency and human-AI collaboration in Industry 5.0.

Future AI Competency and Skill-Gap Analysis for Strengthening AI Talent Development in Undergraduate Education

Kanda Sorn-In — 2026-09-23

Artificial intelligence is reshaping socio-technical systems, requiring universities to realign undergraduate curricula with emerging AI competency demands. This study introduces an integrated analytical framework to identify essential AI competencies and diagnose systemic skill gaps critical for future AI workforce development. Evidence is synthesized from five sources: international competency standards, stakeholder requirements, national AI and digital policies, labor-market intelligence derived from text mining, and global future-skills trends. The analysis delineates a structured set of technical, ethical, and innovation-oriented competencies, while revealing significant gaps in MLOps, large language model engineering, responsible AI governance, and human-centered AI design. The resulting AI Master Competency Taxonomy provides evidence-based guidance for designing industry-aligned, future-ready AI curricula that advance national strategic priorities and promote responsible and sustainable AI development.

Deep Learning–Based Fault Detection and Classification in AC Microgrids using CNNs

Deepak Kumar Lal — 2026-09-23

Since there are issues with bidirectional power flow, fault current levels are extremely low, and operating conditions are dynamic in nature, with high challenges in terms of fault detection and classification with distributed energy resources in AC microgrid. To maintain system reliability and assure promptness in fault detection and response, it is extremely important to properly identify common faults like LG fault, LL fault, LLG fault, and three-phase faults. A new method based on deep learning and a deep learning method based on a convolutional neural network are presented in this paper. With this method, there would not be a need to specifically extract features. Since a large amount of data was simulated in terms of a modified 7-bus AC system created in RTDS software and tested in different fault conditions and operating conditions, a large amount of data was generated. The model finally uses a dense layer and a SoftMax classification. A comparison was made on its performance of the conventional machine learning algorithms such as Multilayer Perceptron Classifier, Decision Tree Classifier, Naïve Bayes Classifier, and the Support Classifier Machine. In the simulation, the performance of the proposed model was found to be superior to the existing methods with an accuracy of up to 99%, thereby clearly distinguishing between symmetry and asymmetry through the analysis of the confusion matrix used in the model.

Real-Time IoT-based Power Generation Forecasting in Microgrid System

Narendra Babu P — 2026-09-29

A microgrid network is a small scale electrical grid network which is used to generate electrical power from various renewable energy resource such as wind energy and solar energy to provide reliable and consistent power to the local customer. Wind energy and Solar energy are both weather dependent renewable resource due to this intermitence power generation occur and unstable quality power generation cause voltage fluctuation between point of common coupling and load in microgrid system. Therefore in order to supply high quality sufficient, consistent, and stable power to critical loads it is mandatory to study on accurate power forcasting in microgrid system. Power outages power in microgrid vary the power consumption and affect electrical tariff and changes human behaviour in response to the changes in electrical tariff. The intermittence and uncertainty renewable energy may cause unstable quality power generation in microgrid. The paper presents the IoT based power forecating in microgrid systems. Moreover, real-time power quality characteristcks are also presented in the paper.

Unbalance Reduction in LV Networks with High PV and EVs via Droop-Controlled BESS

Konepadith PHETSISOUK — 2026-09-29

Low-voltage distribution networks are experiencing increasing phase current unbalance due to the widespread installation of single-phase photovoltaic systems and electric vehicle chargers. These unbalance results in increased power losses, transformer thermal stress, and accelerated asset degradation. This article proposes a decentralized drop control method using three independent single-phase battery energy storage systems mounted on LV transformers. The proposed control scheme regulates the charging and discharging of each BESS based on the deviation between individual phase currents and the average phase current, while explicitly considering the state of charge limitation. This method relies only on local current measurements and does not require centralized communication. The effectiveness of the proposed method was validated using a real residential low-voltage distribution network with high PV and EV usage. Simulation results demonstrate a significant reduction in the current unbalanced factor, improved phase current symmetry, reduced feeder losses, and relief from transformer thermal stress. Technical and economic assessments further demonstrate that the proposed strategy is a more cost-effective alternative to traditional grid strengthening in low-voltage networks.

Data-Driven Intelligence: Machine Learning Models for Short-Term Load Forecasting of a Substation Distribution Transformer

Shilpa AralasuraliSubramanya — 2026-09-23

With the increasing complexity of modern power distribution networks and the rising demand for reliable electricity, accurate load forecasting at the distribution level has become essential for efficient energy management. This study presents a performance-focused analysis of machine learning based short-term load forecasting (STLF) of a substation distribution transformer (SDT). Two STLF methods are proposed: (i) Direct load forecasting method: directly forecasts the SDT load using a dedicated SDT load model; (ii) Feeder aggregation method: forecasts individual feeder load using feeder specific load model and aggregates them with optimized weights to estimate the SDT load. Both methods perform real and reactive power forecasting with a half-hour-ahead prediction horizon and five-minute resolution. The models are trained and tested on data collected from an actual SDT and its connected feeders, ensuring real-world relevance and reliability. The proposed STLF methods offers practical insights for utility operators and system planners by supporting better informed decision-making in short-term load management.

Robust Intrusion Detection with GCN-AE: A One-Class GNN Approach

zahra eskandari — 2026-09-29

Designing Intrusion Detection Systems (IDS) that are robust against both known and emerging threats remains a critical challenge in cybersecurity. Traditional graph-based IDS approaches often rely on supervised learning, which requires large labeled datasets and struggles against evasion tactics that deviate from known attack patterns. To overcome these limitations, recent research has explored unsupervised, self-supervised, and semi-supervised learning methods that focus on modeling normal behavior to detect anomalies. Autoencoders have proven effective in learning compact and informative representations. Building on this, we propose GCN-AE, a novel one-class Graph Neural Network (GNN) model that integrates Graph Convolutional Network (GCN) layers within an Autoencoder framework. The model incorporates a specialized message-passing mechanism designed to capture normative network traffic behavior, enabling it to detect deviations without prior knowledge of attack signatures. We evaluate GCN-AE on the NF-UQ-NIDS-v1 dataset under a binary classification setting. Experimental results show that the proposed model outperforms existing approaches and exhibits strong robustness against zero-day, adversarial, and camouflage attacks. These results highlight the promise of combining GCNs with Autoencoders to enhance the adaptability and resilience of network-based IDS, providing a more effective solution for detecting both known and unknown threats in real network environments.

Degradation-Aware Stochastic BPSO Optimization of Smart Home Energy Management under TOU Tariff

Adool Kruekaew — 2026-09-23

The increasing deployment of photovoltaic (PV) systems, residential energy storage systems (ESS), and electric vehicles (EVs) presents new opportunities for reducing household electricity costs under Time-of-Use (TOU) tariffs. However, effective coordination of ESS charging/discharging, Vehicle-to-Home (V2H) operation, and deferrable load scheduling remains challenging, especially when battery degradation is considered. This paper proposes a degradation-aware Home Energy Management System (HEMS) based on a stochastic Binary Particle Swarm Optimization (BPSO) framework for smart homes equipped with PV–ESS–EV systems. The model integrates a power balance formulation, ESS/EV operational constraints, V2H availability, load-shifting decisions, and a throughput-based degradation cost to optimize 24-hour household energy scheduling. Five operational scenarios (S1–S5) are evaluated to quantify the impacts of ESS integration, V2H, load shifting, and degradation-aware control. The results show that ESS and V2H significantly reduce peak-period grid import, while BPSO-based load shifting achieves the lowest daily operating cost. Incorporating degradation awareness reduces ESS throughput by 30.81%, providing a cost–health trade-off that extends battery longevity. The proposed framework demonstrates an effective and practical approach for improving economic performance and sustainable operation of residential PV–ESS–EV systems under TOU pricing.

Interference mitigation for indoor proximity-based visible light positioning using receiver's angle diversity

Poompat Saengudomlert — 2026-09-29

This work considers proximity-based indoor visible light positioning (VLP) in which the user location is specified through the coordinate of the closest light emitting diode (LED) light source. For applications with human users, tilting of a handheld receiver device may lead to positioning errors due to signal interference from an adjacent LED light source. This work investigates the use of multiple photodiodes (PDs) with angular diversity to improve the positioning accuracy in the presence of receiver's tilting, with a unique focus on using multiple PDs to mitigate the problem of signal interference instead of widening the coverage of each LED. First, it is argued that at least four PDs are required on a pyramid structure to effectively mitigate signal interference. Then, performance analysis in terms of the positioning error probability is presented, to be followed by results on the outage percentage defined as the percentage of receiver positions at which positioning error probabilities exceed some acceptable threshold. After evaluating error performances, the error probability expressions are used to optimize the orientation angle of the PDs attached to a pyramid structure. Finally, hardware experiments are conducted to validate the proposed approach.

Blockchain-based Secure Log Management with LLM-assisted Querying System

Satit Kravenkit — 2026-09-29

This research presents a blockchain-based log management system that integrates an LLM-assisted query mechanism to enhance security and provide contextual insights into storage event logs. The modules for decentralized storage event logs, including registered users, log retrieval and storage, and log auditing, are built on a blockchain deployed via an Ethereum smart contract, which enhances the following. EdDSA-based digital signatures (Ed25519) are applied to audit log events to ensure integrity. Additionally, log files are encrypted using asymmetric encryption (Curve25519) to protect confidentiality, while the decentralized IPFS (InterPlanetary File System) stores log file metadata on the blockchain, reducing storage size. In the querying module, an LLM-based natural-language model is integrated. Moreover, we organize the logs into semantic groups and hourly summary tables for LLM-assisted analysis. Users can analyze log data without complex query rules. The experiment covers four evaluations: First, functional correctness. The storage event log modules correctly support secure log management and verify log audit. Second, IPFS delivers practical performance with average upload times (0.41 s) and download times (0.19 s). Third, blockchain operations require reasonable gas usage. Fourth, the Llama 3.1 (8B) achieves the best querying performance, with Correctness (0.80), Faithfulness (0.53), and Helpfulness (0.62) supporting practical log analysis.