Electricity is an important commodity in modern society. All modern households, commercial industries and critical infrastructure requires electrical energy supply to function effectively. As a result, electricity has become an essential resource for economic growth, technological development and social progress.
Delivering electrical energy to consumers requires complex infrastructure and systematic processes. Electricity is produced in distant generating power plants using various energy sources. From power plants, it is transmitted through high-voltage transmission lines although dangerous the power transfer efficiency is significant. It is then reduced to lower voltage and distributed to the consumers. The interconnected structure that enables this entire process is known as the electrical power grid.
The electrical power grid is considered one of the most sophisticated engineering systems ever developed. It consists of numerous devices, facilities and plants including a central command that controls and monitors the system. These components must operate together in a coordinated manner to ensure reliable electricity delivery. Over the years, significant improvements have been introduced in power system design, operation, protection, and control. These advancements have contributed to higher reliability, improved efficiency, and greater system stability. Nevertheless, the power grid continues to face new challenges as electrical demand increases and system requirements become more complex.
The growing consumption of electricity has created a need for further improvements in power system operation and planning. At the same time, modern power systems are integrating larger amounts of renewable energy resources. These developments introduce additional operational uncertainties and technical challenges. Engineers must therefore develop more precise methods for monitoring, controlling, and optimizing power system performance. Continuous improvement is necessary to maintain reliability and support future energy requirements.
Despite the need for innovation, testing new strategies directly on actual power systems can be difficult and potentially dangerous. Electrical grids operate under strict reliability and safety requirements. Any unsuccessful modification may lead to equipment damage, power interruptions, financial losses, or safety hazards. Furthermore, conducting large-scale experiments on operational power systems can be expensive and time-consuming. These limitations make it impractical to evaluate every proposed improvement through direct implementation. Consequently, researchers and engineers often rely on simulation techniques to investigate system behavior and assess the feasibility of new approaches.
Simulation has become an essential tool in modern power system research. It provides a controlled environment in which different operating conditions can be studied without affecting real-world infrastructure. Researchers can analyze system performance, evaluate control strategies, investigate fault conditions, and predict system responses under various scenarios. The continuous advancement of computational capacity has significantly enhanced the capabilities of simulation tools. Modern computers can process large datasets and perform complex calculations with high speed and accuracy. These capabilities allow researchers to construct detailed mathematical models that closely represent empirical system behavior. The success of simulation-based research has also been supported by the availability of standardized test systems and modeling frameworks. Many of these resources are documented in publications and repositories in the Institute of Electrical and Electronics Engineers (IEEE). Such resources provide researchers with reliable platforms for validating methodologies and comparing results across different studies. Thus, simulation has become a widely accepted approach for evaluating proposed solutions before their potential implementation in real-world power systems.
This reviews five research studies that focus on the development and assessment of different techniques for improving power system performance. The selected studies explore various modeling approaches, optimization methods, and combinations of analytical techniques. Each proposed method is evaluated through simulation-based analysis. Mathematical formulations are used to establish theoretical foundations, while statistical measures are employed to assess performance and effectiveness. Through a comparative examination of these studies, this paper seeks to determine the level of success achieved by the proposed models. The findings may provide valuable insights into current developments in power system research and contribute to a better understanding of emerging approaches for electrical grid improvement.
A Boundary-Compensated Partition-Based Parallel Graph Neural Network for Weak-Bus Identification in Interconnected Power Grids
The study highlights the limitations of two methods, namely conventional full-graph neural networks for large-scale power grids and direct graph partitioning techniques used for identifying weak buses in the system. Boundary regions that are not covered by the model may become vulnerable, and their susceptibility may remain undetected. To address this limitation, the paper proposes a boundary-compensated partition-based parallel graph neural network framework for the identification of weak buses. The IEEE 57-bus benchmark, together with mechanism-based node and branch vulnerability labels, was used to evaluate the proposed method. The analytical framework successfully captures weak buses within the power system. Furthermore, local aggregation, boundary transmission, and corridor-driven vulnerability propagation were identified as key indicators of weak buses. However, the study limits its validation to a medium-scale benchmark rather than demonstrating scalability across large-scale power systems.
A Novel Hybrid Platform Based on Deep Learning for Fault Detection and Localization in a Smart Distribution Grid
Rapid and accurate fault detection and localization are essential for ensuring the reliability of smart transmission systems. Traveling-wave (TW) methods provide high fault-location accuracy under strict operating conditions, whereas data-driven models are generally more robust but often exhibit lower precision. The paper proposes a unified and modular framework that integrates the TW analytical method with a one-dimensional convolutional neural network (1D-CNN) model, allowing the strengths of both approaches to complement each other. The key contribution of this study is an adaptive decision strategy that combines the outputs of the TW method and the 1D-CNN model to generate a more accurate and reliable fault-location estimate. The framework’s performance is systematically evaluated using standard detection and localization metrics. The hybrid methodology enhances both accuracy and robustness, demonstrating its potential as a solution for advanced monitoring and protection in modern power grids.
Data-Driven Physics-Informed LSTM for Voltage Regulation in Active Distribution Networks
The integration of renewable energy sources, particularly photovoltaic (PV) generation, into the power grid creates challenges in voltage regulation due to the distributed nature of these energy sources. Several control methods have been developed, but each has its own limitations. Droop control operates locally, central optimal power flow requires full network observability and control, and multi-agent deep reinforcement learning (MADRL) methods involve long training times and significant algorithmic complexity. The paper proposes the Optimal Historical Selection and Forecasting (OHSF) scheme, which combines a physics-informed long short-term memory (LSTM) network with an online sensitivity-based correction loop for medium-voltage active distribution networks. The results demonstrate reduced average voltage deviations across all PV buses in simulations of modified IEEE 33-bus and 69-bus test systems, while also achieving significantly shorter training times.
MISSA-BPNN-Based Surrogate Model for Wind-Induced Stress Prediction in Vulnerable Regions of Transmission Towers
Stress monitoring of transmission towers under strong wind conditions remains challenging. Conventional contact-based sensors are complex to install, difficult to maintain, and prone to damage. Recent advances in non-contact displacement measurement technologies, such as laser measurement and machine vision, have created new opportunities for structural monitoring. The paper proposes a wind-induced stress surrogate model based on a Multi-Strategy Improved Sparrow Search Algorithm and a Backpropagation Neural Network (MISSA-BPNN) to identify structurally vulnerable regions of transmission towers. The results demonstrate the successful identification of vulnerable regions in the transmission towers analyzed. This method provides a new approach for monitoring transmission line infrastructure.
Short circuit current limitation using series reactors in 20 kV distribution feeder
Short-circuit faults produce extremely high fault currents that can cause significant damage to electrical equipment. These faults can be severe and may significantly reduce system reliability. One method of limiting fault-related damage is the installation of a series reactor, also known as a current-limiting reactor (CLR). The study uses ETAP simulation software based on IEC 60909 standards to evaluate the effects of four types of short-circuit faults: three-phase faults, single-line-to-ground faults, line-to-line faults, and double-line-to-ground faults. The results confirm that series reactors can significantly reduce fault currents, thereby minimizing potential damage to electrical equipment.
Conclusion
The five recently published research articles focus on different aspects of the power grid. The first article focuses on system vulnerability, the second presents a new fault-detection method, the third addresses voltage regulation challenges associated with renewable energy integration, the fourth examines the structural assessment of transmission towers, and the fifth discusses the protection of power equipment and transmission lines against faults and their rapid recovery. Collectively, these studies complement one another and contribute to the development of a more robust and stable power grid.
References
Qin, J., Zhang, Z., Li, F., Xue, Y., Si, Y., & Su, L. (2026). A Boundary-Compensated Partition-Based parallel graph neural network for Weak-Bus identification in interconnected power grids. Energies, 19(11), 2630. https://doi.org/10.3390/en19112630
Souhe, F. G. Y., Ekemb, G., Mbey, C. F., Kakeu, V. J. F., & Boum, A. T. (2026). A novel hybrid platform based on deep learning for fault detection and localization in a smart distribution grid. Journal of Electrical and Computer Engineering, 2026(1). https://doi.org/10.1155/jece/8886698
Hein, H., Yu, H., Yu, L., & Deng, Z. (2026). Data-Driven Physics-Informed LSTM for voltage regulation in active distribution networks. Energies, 19(11), 2609. https://doi.org/10.3390/en19112609
Wang, F., Zhang, T., Tang, Y., & Liu, Y. (2026). MISSA-BPNN-Based Surrogate Model for Wind-Induced Stress Prediction in Vulnerable Regions of Transmission towers. Processes, 14(11), 1785. https://doi.org/10.3390/pr14111785
Siregar, R. H., Farras, A., & Syahrizal, S. (2026). Short circuit current limitation using series reactors in 20 kV distribution feeder. Journal Geuthee of Engineering and Energy (JOGE), 5(1), 62–74. https://doi.org/10.52626/joge.v5i1.94