https://ijrtei.sbu.ac.ir/ International Journal of Research and Technology in Electrical Industry en daily 1 Thu, 01 Jan 2026 00:00:00 +0330 Thu, 01 Jan 2026 00:00:00 +0330 https://ijrtei.sbu.ac.ir/article_107390.html The increasing integration of renewable energy resources and digital communication technologies has significantly improved the intelligence and operational flexibility of modern smart grids. However, the growing reliance on automation cyber–physical systems has also increased vulnerability to cyberattacks, threatening the reliability and security of energy management processes. Interconnected local energy networks (ILENs), which enable coordinated operation of distributed energy resources, are particularly exposed due to their dependence on sensing, communication, and control infrastructures. This paper proposes an AI-powered cyber-secure framework for robust energy and flexibility scheduling in ILENs. A robust bi-level optimization model is developed, where the upper-level operator coordinates energy and flexibility exchanges among interconnected networks, while lower-level operators optimize local energy management under uncertainty. Robust optimization is incorporated to address uncertainties in renewable generation, load demand, and electricity prices, thereby enhancing scheduling resilience. To assess cybersecurity performance, false data injection (FDI) attacks are modeled using a Gaussian probability distribution. An XGBoost-based anomaly detection and correction mechanism is employed to identify compromised measurements and recover corrupted data before scheduling decisions are executed. Simulation results demonstrate a maximum detection accuracy of 91.67%, while the proposed correction strategy effectively restores system performance to near-normal operating conditions. The proposed framework simultaneously strengthens operational robustness and cyber resilience, offering a secure and reliable solution for future automation-enabled smart grids. https://ijrtei.sbu.ac.ir/article_107401.html The widespread application of capacitor banks in transmission substations has increased the risk of harmonic resonance under different network operating conditions. Accurate identification of resonance-prone operating states is therefore essential for maintaining power quality and system reliability. This paper presents a practical methodology for harmonic resonance assessment based on frequency sweep analysis and the Resonance Risk Index (RRI). First, the resonance frequencies of different capacitor-bank operating states are determined using a detailed DIgSILENT PowerFactory model of a real transmission network. The RRI is then calculated by considering the resonance frequency, harmonic distortion level, and network short-circuit capacity. Unlike conventional approaches that evaluate resonance risk only under normal operating conditions, the proposed method further investigates the influence of transmission-line outage, transformer outage, and 20-kV bus-coupler opening on resonance risk. The results demonstrate that network topology significantly affects both the resonance frequency and the corresponding RRI. While some switching operations increase resonance risk, others reduce it by shifting the resonance frequency away from dominant harmonic orders. These findings indicate that resonance risk is a dynamic characteristic rather than a fixed network property and highlight the importance of considering practical operating conditions in harmonic resonance assessment. The proposed approach provides transmission system operators with a practical decision-support tool for identifying high-risk operating states and selecting appropriate operational or technical mitigation measures. https://ijrtei.sbu.ac.ir/article_107350.html Data-driven methods for DSA, particularly those based on deep learning (DL), have shown promising results. However, these approaches face challenges such as topology modifications and missing data, which hinder their implementation in real-world power systems. On the other hand, most existing DSA methods function as black-box models, offering limited interpretability and failing to provide actionable insights or recommended corrective measures to prevent power system instability. This paper proposes a multitask transfer learning framework to address these challenges. In this framework, three separate models are trained using a multitask learning approach. Each model is responsible for a specific task, including missing data reconstruction, DSA, and intelligent load shedding (LS). Generative adversarial networks, an unsupervised DL technique based on the competition between two neural networks, are employed to handle missing data. For DSA and LS, graph convolutional networks are employed to enhance accuracy by effectively capturing the topological structure of the power system. The multitask learning framework for data reconstruction, DSA, and LS is well-trained for a specific topology; however, it experiences a decrease in performance accuracy when applied to unseen topologies. To address this challenge, transfer learning is utilized to update the model with a small dataset. Implementing the proposed framework on the IEEE 39-bus system shows its positive performance in solving the challenge of missing data and topology changes.